B.2 Executive Summary

Size: px
Start display at page:

Download "B.2 Executive Summary"

Transcription

1

2 B.2 Executive Summary As demonstrated in Section A, Compute Canada (CC) supports a vibrant community of researchers spanning all disciplines and regions in Canada. Providing access to world- class infrastructure and expert personnel supports Canadian researchers. All Canadian university researchers have equal opportunity to access the CC resources. Larger requests are accommodated through an annual peer review allocation process that ensures Compute Canada is providing access and support to the most promising research in Canada. The advanced research computing (ARC) needs of the Canadian research community continue to grow as the next generation of scientific instruments is deployed, as ARC becomes relevant to answering key questions in an ever broader list of disciplines, as new datasets are gathered and mined in innovative ways, and as technological advances allow researchers to construct ever more precise models of the world around us. The current CC infrastructure must keep pace with the needs of Canadian researchers. This proposal addresses the urgent requirement to replace many aging systems with a consolidated set of systems designed expressly to meet Canadian research needs. These systems are designed to balance the need for technical innovation, with ongoing productivity, avoiding technologies that may require many months of refinement before research groups can effectively use them. The new systems are designed to meet the needs of the broad range of users identified in CC s Strategic Plan. These upgrades will improve services to both traditional users who focus on the number of cores available, and newer users who need a balance of technology leadership as well as service and support leadership. In order to promote effective and efficient use of new infrastructure, CC will offer researchers common identity management, software environments and data management tools across a national network of facilities. Integrated services will be matched with the development of a nationally coordinated support regime. Local user support will continue to be provided by on- campus personnel, augmented by a national network of subject matter experts as well as supported user communities. As the CC data centre footprint is consolidated, a stronger network of systems administrators will be able to serve a wider range of systems, both locally and remotely. Working with our regional partners, we will create a deeper pool of expertise in critical areas such as file systems management, networking, systems software, applications software, and code optimization. This will allow CC to increase the level and professionalism of its service to the community without significantly increasing investments in personnel. Compute Canada, through consultation with Canadian researchers, has developed a well- documented forecast of needs versus the current capacity and the expected capacity with the current planned investments. The funding available through the Canada Foundation for Innovation s (CFI s) Challenge 2, Stage- 1 Cyberinfrastructure Initiative is not sufficient to meet all of these needs. As such, choices must be made about which needs will be supported, and to what degree. CC has developed a balanced approach as the recommended baseline option in this proposal. Two alternative options have been developed which shift the balance in favour of either tightly coupled computations or data analytics. Pursuing these alternative options in stage- 1 comes at a cost to the existing CC supported science programme. Assuming the baseline option is chosen in stage- 1, the alternative options are directions Compute Canada is likely to pursue with the additional funding available in stage- 2. The technological refresh and changes to the service delivery model in stage- 1 will empower Canadian researchers to pursue leading- edge research. The extensive benefits to Canada documented in Section A will continue as Canadians continue to push forward the boundaries of their disciplines and compete on an international stage. Revolutionary change in many fields and the resulting societal benefits now rely critically on ARC. From personalized medicine to better aircraft design, from the modelling of novel materials to modelling the Canadian economy, CC will continue to enable the creation of new knowledge across a broad range of domains. 40

3 B.3 Need for the Infrastructure B3.1 Immediate and Pressing Needs CC currently operates 50 systems in 27 data centres across the country. More than half of the roughly 200,000 computational cores in operation today were deployed in 2010 or earlier and are hence already beyond their normal lifespan of five years. These pre systems also provide more than 25% of currently available storage resources. The vast majority of the remaining resources were deployed in 2011 and 2012 and will reach their nominal lifespan in 2016 or As it stands today, most of the pre systems are on limited maintenance contracts covering only critical components. For the sake of system reliability, there is an urgent need to replace existing infrastructure. Ignoring concerns about reliability, operating costs for maintenance and repairs are growing yearly as older systems reach the end of their originally purchased warranties and manufacturers no longer offer service on obsolete components. In addition, normal improvements in efficiency mean that modern systems would deliver similar computational performance for much lower electrical energy costs. Maintenance and energy costs need to be reduced to allow increased investment in support and service. Finally, regardless of reliability or the cost of operations, CC has reached the limits of compute and storage capacity that can be allocated to its most excellent research users. Demand continues to increase, while the ability to meet that demand is falling. B3.2 Responding to the Needs of Existing and Emerging Research Communities The needs of the research community have evolved since the last round of major capital purchases by CC. The rapid growth in data- intensive research has strained the capacity of CC to meet data storage needs for ongoing research projects. The problems being solved via modelling of materials, biological molecules, and other complex systems (e.g. earth- ocean) have increased in precision and concomitant computational intensity. Adoption of accelerators (GPUs) is revolutionizing certain types of problem solving, such as machine learning (so- called deep learning). For some emerging areas (e.g. image analysis), the required system memory per computational core has exceeded the capacity of most existing CC systems, such that use of these systems is becoming less efficient for some problems and impossible for others. In addition to hardware infrastructure changes, the way that researchers interact with the infrastructure has also changed dramatically in the last five years with the emergence of cloud computing and the proliferation of scientific gateways and data portals. In order to adapt to modern workloads, there is an urgent need to replace existing infrastructure. As illustrated in Section A, CC now serves a rapidly growing number of researchers across a wide range of disciplines. Assessing the ARC needs of such a broad group is challenging and CC has undertaken extensive consultations in order to engage the community. This consultation has included: A needs survey distributed to all CC users in the autumn of 2013 (more than 200 faculty responses). More than 20 in- person consultations at various Canadian campuses in the winter of This was associated with the writing of the attached Compute Canada strategic plan. Several online- only consultation sessions were also offered. A call for white papers was issued in summer papers were received from a variety of disciplinary bodies and institutions. Advisory Council on Research (ACOR) was formed in 2013 and met regularly through proposal submission to give input to the planning process. A draft infrastructure proposal was posted on the Compute Canada website and was broadcast to the CC researcher mailing list (more than 10,000 people) in January

4 In- person consultations were held at 6 locations across Canada in January This was followed by an online- only consultation session. In addition, user data from the Compute Canada Database (CCDB) was mined for the period to search for usage trends. Existing usage data was then combined with the consultation data described above and was compared to international trends. While CC has made extensive efforts to capture needs from all areas of science, there remains an unavoidable bias towards existing CC users compared to researchers in emerging disciplines due to the different response rates from the two communities. B.3.3 Current and Anticipated Needs by Thematic Research Area Each of the thematic research areas identified in Section A will see increasing demand for infrastructure over the next 5 years. In some cases, this is due to a constant progression of the field towards more complex models and more compute- intensive treatments. In other cases, anticipated advances in instrumentation are expected to drive data- intensive research in a certain field. Some examples are provided below, organized by the thematic areas of Section A. Common to all thematic areas is the need for expert personnel to enable efficient use of ARC resources in cutting- edge research. Theme 1: Materials Science, Condensed Matter and Nanoscience A white paper in this area was submitted to the CC SPARC process by 28 faculty members from 12 Canadian universities. That paper illustrated that the growth in this field is driven by the need for realistic and experimentally relevant real materials simulations. Materials are studied on multiple length and timescales and the methods vary according to those scales. Much of the computation is accomplished today using homemade codes specialized to solve a certain problem of interest. However, Canadians are also involved in some large multi- national initiatives to produce more general- purpose software. The United States is currently funding the Materials Genome Initiative to speed up our understanding of the fundamentals of material science, providing a wealth of practical information that entrepreneurs and innovators will be able to use to develop new products and processes. In particular, The initiative funds the development of computational tools, software, new methods for material characterization, and the development of open standards and databases. As such, this area is poised for substantial growth in computational need (at least a factor of 5 in the next 5 years). Roughly half of the usage in materials science is expected to be serial in nature while the other half would benefit from being able to run parallel codes on highly connected machines. Given the choice, this community would maximize the number of cores deployed over optimization of machine interconnect. The importance of acceleration via FPGPU and GPGPU is evolving rapidly. Theme 2: Chemistry, Biochemistry and Biophysics This area currently represents the single largest utilization of CC CPU by discipline. This CPU is used to solve problems using molecular dynamics (MD) simulations, quantum mechanical calculations that explore electronic and molecular structure, ab initio MD simulations that derive molecular interactions from first principles, and hybrid techniques. In order to achieve further advances or to provide new insights, researchers need to move to more detailed descriptions and better models, larger systems, and/or longer timescales. Given that these approaches are in essentially all cases computationally intensive, this translates into significant need for greater computational power, with implications such as increased memory, increased storage, and increased parallelism (need for fast interconnects). Approximately 65% of the CPU time consumed by computational chemistry calculations on CC resources today is by jobs which are at least moderately parallel (64 cores) and 12% is consumed by highly parallel jobs (at least 1024 cores). This community is also extensively exploring the use of GPU accelerators and sees at least a factor- of- four improvement in calculation speed when supported by an accelerator. 42

5 Theme 3: Bioinformatics and Medicine Over the next decade almost every biomedical investigation in basic and clinical research will be enabled through characterization of an accompanying genome sequence. Genomic technologies have become a critical component not only in human health research but also in other fields such as: agriculture, fisheries, forestry and mining. With next- generation sequencing technologies revolutionizing the life sciences, data processing and interpretation, rather than data production, has become the major limiting factor for new discoveries. In this context, the availability of advanced research computing resources has become a key issue for the genomics community. Advanced Research Computing Resources and Needs at 4 Canadian Genome centres (submitted to SPARC process) The increased demand in genomics will be primarily driven by three factors: improvements in instrumentation, the use of more advanced analysis strategies on acquired genome data, and increased demand for access to informatics infrastructure to utilize large international public datasets. The estimated growth in this area is at least a factor 8 in CPU and nearly a factor of 30 in disk storage over the next 5 years. Generally speaking, computations in this area require a Big Data infrastructure including high- throughput disk arrays. For some types of analysis, high- memory nodes are required (e.g. at least 512GB per node). Most applications do not take advantage of a high degree of parallelism. Data privacy restrictions are important considerations in serving the ARC needs in this area. Many projects involve identifiable personal health information that must be protected by both appropriate policies and appropriate technological safeguards. Medical research is now the largest category of special resource allocation requests received by Compute Canada each year. While the number of requests is growing rapidly, each request is not (yet) as compute or storage intensive as requests from some other disciplines. Adopting a better security posture at new data centres is an important adjustment that CC must make in order to serve this community. Since 2012, CC has added two major centres (BC Genome Science Centre and HPC4Health) to the organization in this area. In 2015, CC has become a partner in a successful Genomics Innovation Network proposal to Genome Canada and is generally playing an active role in supporting the Canadian genomics community. Providing service to this community is a clear priority for CC and can only be enabled through new infrastructure purchases. Theme 4: Earth, Ocean and Atmospheric Sciences A white paper on the needs of the ocean modelling community from researchers at 10 Canadian universities was submitted to the Compute Canada SPARC process. This community strives to improve our basic understanding of oceanographic processes and our ability to simulate, predict and project physical, biological and chemical ocean characteristics on timescales from days, weeks and seasons to centuries. This community currently uses parallel codes which scale well in the range from cores and so requires large compute clusters with high- speed interconnect between the nodes. The lack of a dedicated large parallel machine in Compute Canada with scheduling optimized for large jobs means that members of this community typically wait for days to begin a single calculation. The presently available infrastructure limits the temporal and spatial resolution possible. Doubling the resolution leads to an increase in required compute power of roughly an order of magnitude. Moving from 2- dimensional to 3- dimensional models, which are now becoming more common, increases the required computational power by 2-3 orders of magnitude. This community requires increased capacity in tightly coupled cores in order to remain competitive. Theme 5: Subatomic Physics and Astronomy The Canadian subatomic physics community is involved in several high- profile global experiments with significant computational, storage and advanced networking needs. A group of 39 Canadian faculty members currently participate in the ATLAS experiment at the Large Hadron Collider (LHC). Run I at the LHC completed in 2012 and featured the discovery of the Higgs boson. Run II begins in the summer of 2015 with upgraded energy 43

6 and a doubling in the data- taking rate. The demand for high- throughput storage will grow throughout Run II, which ends in mid The instrument will then undergo upgrades and will return in the early 2020s at an even higher data- taking rate. Several other major subatomic experiments served by Compute Canada are also being upgraded or are coming online in the next 5 years. ATLAS compute and storage needs in Canada are currently met by the Tier- 1 computing centre at TRIUMF and by four Tier- 2 computing centres within Compute Canada. In preparation for this proposal, Compute Canada and TRIUMF have agreed to pursue a partnership in which the current TRIUMF Tier- 1 staff would join Compute Canada and Tier- 1 functionality would be transitioned from TRIUMF to one of the new consolidated Compute Canada data centres. The Tier- 1, which requires 24x7 support and a high- bandwidth connection to CERN, would be co- located with a Compute Canada Tier- 2 centre. As part of this process, Compute Canada would consolidate ATLAS Tier- 2 support from four sites to two. This is a more efficient operational arrangement and represents a major redesign for ATLAS computing support in Canada. Experimental subatomic physics requires large quantities of high- throughput storage and nearby computation cores to process the data. The jobs are generally serial, or parallel over a small number of cores (e.g. 8), though GPUs are starting to be used and provide a significant advantage for specific types of calculations. Memory requirements are generally moderate (e.g. 4GB/core). In future, centres that support ATLAS must provide 100Gb connectivity to the LHCONE network. Theoretical subatomic physics often relies on parallel codes scaling on interconnected nodes into at least cores, depending on the sub- discipline. CANFAR, a collaborative effort of the Canadian university astronomy community, currently makes Canadian astronomy data available to researchers around the world. This platform also provides compute resources that enable those researchers to process and analyze that data. The CANFAR platform operates on Compute Canada resources. The Canadian Astronomy Data Centre (CADC) currently hosts copies of the raw data, as well as database and other support services that are necessary for the proper functioning of CANFAR. Compute Canada and CADC are currently discussing a 3- year plan to migrate these core services to Compute Canada (costs to be paid by the National Research Council, outside the scope of the MSI project award). The Compute Canada services would continue to be supported by CADC personnel. The CANFAR platform has recently been migrated from a Nimbus cloud to the new Compute Canada cloud systems, which run OpenStack. For some image processing, for example, it requires high- memory nodes (512GB per node). While observational data processing tends to be serial in nature, this is not the case for theoretical astronomy, astrophysics and astrochemistry. These calculations require a large number of computational cores in tightly coupled systems. Theme 6: Computer and Information Sciences Computer scientists naturally push some of the technological boundaries of ARC in a variety of technical domains. Compute Canada serves a diverse set of Canadian computer scientists including a strong machine learning community. In particular, the Canadian machine learning community is making extensive use of GPU co- processing in order to mine data using deep learning techniques. These techniques are relied upon for the artificial intelligence behind modern image and speech recognition and are expected to see significant growth in breadth of application. In the coming years, the group of Yoshua Bengio expects to require 240 GPUs for his 60- person laboratory. Across Compute Canada, this research field alone could use productively more than 1000 GPUs, which offer 10-20x speed- ups compared to conventional CPU processing for this type of application. Theme 7: Social Sciences and Humanities While Compute Canada resource usage in social sciences and humanities is currently small as a fraction of overall compute and storage usage, this is a growth area in which the delivery and support of services is often more important than the scale. One limiting factor in the exploitation of CC resources by researchers in the social sciences has been the need to manage private data sets. While CC has recently taken responsibility for housing and managing RCMP crime data 44

7 at a particular site in collaboration with a local computational criminology group, this is an exception rather than the norm. Adopting an enhanced security posture (both in policy and technology) is vital to supporting social science researchers. Over the last year, CC has engaged in detailed discussions with the Canadian Research Data Centre Networks (CRDCN) and Statistics Canada around access by researchers to Statistics Canada datasets. CC is assisting with the design of the refresh of CRDCN platform and may come to play an ongoing role in this area. CC received a white paper submission from the Canadian Society for Digital Humanities, which laid out their most pressing needs going forward. In addition to enhanced training resources and specialist Digital Humanities (DH) support personnel, they requested a cloud- based web- accessible infrastructure backed by significant storage resources. CC has invited DH researchers to be beta testers of the Compute Canada cloud and is working closely with these groups to ensure that the required cloud services are available on the infrastructure deployed as a result of this proposal. B.3.4 Projecting Demand for Compute and Storage Resources Based on responses to community consultations and analysis of existing usage data, CC has undertaken an exercise to project future infrastructure needs for the Canadian community. The projections below are based on the growing needs of existing Compute Canada users and do not account for anticipated growth in the CC user base. Computation In response to a survey distributed to CC users in fall 2013, computational resources were ranked as their number 1 current and future need from Compute Canada. The SPARC white papers demonstrated a broad need for increased computational resources over the next 5 years as shown in the table below. White Paper Numerical Relativity Subatomic Physics Materials Research Canadian Genome Centres Canadian Astronomical Society Theoretical Chemistry Predicted Increase from Current to x 3x 5x 8x 10x 12x Weighting by current usage by discipline, this leads to an average expected increase of 7x over 5 years. It should be noted that, in some cases, the range of responses within a discipline may include researchers who need 100x over the next 5 years. Based on this and on international norms, the growth rate used here should be considered as a lower bound. Storage Many communities see storage growth rates at least commensurate with their compute growth. However, research communities analyzing datasets collected from a variety of different instruments or agencies see additional storage growth beyond their ability to grow computational power. CC has already witnessed a rapid increase in storage demand that has outstripped the supply at existing sites. The Canadian subatomic physics community has some of the largest storage allocations on Compute Canada resources today. This discipline represents traditional big data. The long timelines of the associated experiments and relative maturity of the field mean that the storage growth rate is predictable and controlled. This provides us with an example of a large base experiencing only modest growth. By contrast, in some 45

8 disciplines the pace of change is very rapid, making it impossible to apply predictable growth limits to the data in advance. As an example, sequencing production in the four largest Canadian genome centres currently doubles every 12 months. The table below illustrates anticipated storage growth from these two Canadian Big Data communities. The growth in disk needs for subatomic physics is a relatively modest factor of 3 over the 5- year period from In contrast, the disk need in genomics increases by a factor of 27 over the same period. Storage Requirements Growth Subatomic Physics Disk (PB) Genome Centre Disk (PB) Total Disk (PB) Subatomic Physics Tape (PB) Genome Centre Tape (PB) Total Tape (PB) In addition, other communities report very rapid growth rates. Neuroimaging researchers supported by a CC Research Platform and Portals award have projected a 14x growth in storage need over the next 3 years. As a result of these expected increases, CC has conservatively assumed an average growth rate of 15x over the next 5 years. Compute and Storage Projections Using the compute and storage numbers above, CC has produced the growth curves shown below. For the compute projections, the unit core- years (CY) is used. This represents the amount of computation that can be performed by a single computational core running constantly for 1 year, or the computations performed by 12 such cores in one month, etc. (based on the cores deployed in the current CC fleet). The solid line represents demand as extracted from recent CC resource allocation competition data. Future years are calculated using the weighted average 7x growth rate over 5 years described above and assuming that the growth is exponential in form. For the supply curve (blue), it is assumed that the full $15M CFI award in 2015 is allotted to Compute Canada, that the baseline option in this proposal is funded and that the resulting equipment comes online in When this comes online, pre systems are assumed to be decommissioned, leading to a net drop in core- count. It is further assumed that the full $15M CFI award in 2016 is allotted to Compute Canada and that this equipment comes online in This leads to the first real increase in core- count since Since there are no further CFI competitions approved at this time, no increases are assumed beyond

9 For the storage projections the unit petabytes (PB) is used. The solid yellow line is again demand extracted from recent resource allocation competitions and the future demand projections (dashed) use the 15x growth rate over 5 years assuming an exponential form. In estimating the supply (blue), the full stage- 1 and stage- 2 Cyberinfrastructure funding is assumed. It is further assumed that some storage from stage- 1 is front- loaded into the 2016 fiscal year in order to meet pressing current demand. B.3.5 Current Job Size Distribution CC currently supports a wide range of computational needs. The figures below provide two ways to view the number of cores used in a typical Compute Canada computation (or job ). The plot on the left shows the number of core years used in CC as a function of the year. The various colours illustrate the fractions of those core years in bins of cores- per- job. It shows, for example, that nearly 50% of CPU consumption in 2014 was by jobs using at least 128 cores. The plot on the right illustrates what fraction of the CC user base (counting project 47

10 groups, not individual users) have submitted at least one job using a given number of cores, shown as a function of time. Further information about parallelism in the CC user community is visible in the table below, which summarizes usage data for In this table, the first column represents the minimum number of computational cores used in a single job. The second column represents the fraction of project groups that have submitted at least one job of at least that many cores. The third column represents the fraction of total CPU usage represented by jobs of at least that many cores. This means, for example, that 19% of user groups submitted at least one job of at least 256 parallel cores and that these jobs represent 31% of all CPU resources consumed in Summary of Data Usage Min. Number of Cores/Job Fraction of Groups (%) Fraction of CPU Usage (%) It should be noted that the size and configuration of CC s current systems limits the ability of Canadian researchers to submit jobs at the largest scales and so has likely limited the growth of the highly parallel bins. To illustrate this effect, consider the SOSCIP BlueGene system that offers service to southern Ontario researchers. This system provided more than 32,000 core- years of computation in 2014 to jobs using at least 1024 cores. Some of these users have shifted their computational workloads from CC systems to the SOSCIP system in order to take advantage of the highly parallel architecture. Others, notably users from the astrophysics community, have found ways to access resources in other countries, including XSEDE in the US and even Tihane- 2 in China. B.4 Efficient and Effective Operation The current distribution of CC data centres and systems reflects the distribution of resources from the seven pre- existing regional consortia that joined to form Compute Canada in Future hardware investment will be optimized on a national level into fewer, larger systems with national service roles. CC expects the current fleet of 27 data centres to be reduced to 5-10 by By concentrating investment in this way, important advantages will be realized: The CC management regime and role will shift, such that the central organization provides oversight for quality control, central processes for configuration change management and security, and coordinated planning for technology refresh. 48

11 Some expert personnel will support enhanced services available across Canada rather than distinct hardware systems. The complexity of the CC enterprise will be reduced by not maintaining 27 bilateral hosting arrangements. Many researchers will no longer need to have their resource allocations split across multiple systems. This eases the burden on research groups. At the same time, it simplifies scheduling and storage allocation procedures for CC. Having a mix of hardware types in a single site is particularly valuable to those groups who require a mix of job types throughout their overall workflow. Better efficiency of operation and economy of scale will be attained by purchasing fewer, larger systems and having fewer support contracts. CC will be aligned with other national and multinational ARC consortia, by heading towards a more sustainable model of operation where hardware resources are centralized at locations where operational conditions are favourable and where qualified on- site staff are available. Access by users and most support staff is via the national wide- area network The purchase of new infrastructure and consolidation of compute centres provides a unique opportunity to rethink both the way CC resources are managed and the way researchers interact with those resources. It will help Compute Canada evolve from today s federation of systems and support into national- level cyberinfrastructure, with support that transcends site and regional boundaries. During the stage- 1 technology refresh, four new sites will receive four new systems (described below), and a number of other systems will be defunded and removed from the CC allocations process. This shift in resources creates an opportunity for a shift in roles and expectations for CC s staff members. Rather than having the majority of services for systems based at the host institution, the future will see support coming from across all of Compute Canada. The on- site support that users value will continue as a key component of Compute Canada s services, and will be augmented by experts from across the nation. A range of activities, from software licensing to 24x7 monitoring and response, will shift from an institutional model to a pan- CC model. CC s leadership, working closely with regional leaders and member sites, will guide personnel towards thinking more broadly about their roles. Personnel will have the opportunity to become increasingly specialized, knowing that their knowledge might be called upon from any CC user at any site. Canada is ideally positioned to become a world leader in national- level support for ARC. Canada has an outstanding research network backbone, a broad mix of research universities, and a strong record of collaborative scholarship. The multi- year shift from having ARC resources plus personnel at member sites, towards centralization of resources while retaining on- site personnel, provides two key opportunities: 1. To pursue an active technology refresh program, in which a limited number of sites host large- scale ARC systems to serve all CC constituencies; 2. To create a pan- Canadian support structure for ARC users, in which on- site talent is augmented by experts from across all member institutions. CC s plans in each area are described in this section. B.4.1 National Centres Four sites have been identified for hosting the next Compute Canada systems, which are anticipated to be available for use by mid All current CC centres, while part of a national network of systems, have traditionally operated with a large degree of autonomy. As an example, all CC researchers currently have equal access to every system in the network, but there is no mechanism to grant administrator privilege at a given site to staff from outside that site. 49

12 Compute Canada has recently established some core principles that define a national site. These core principles were mandatory hosting conditions in the site selection process described later in this document. These core principles are part of the signed agreements between newly selected hosting sites and CC: Allocation of resources on the hosted system(s) will be performed through the Compute Canada resource allocation process. No institution or region will receive preferential access to those system(s). Decisions on hardware procurement will be made through a national process. Local purchasing rules must allow Compute Canada staff to participate fully in the hardware vendor selection process. The host institution will own the purchased system(s). Sites will participate fully in collection and reporting of information about the purchased system(s) operation in accordance with Compute Canada policies. This includes automatic collection of usage information, system up- times, etc. This information will be used to ensure consistent configuration and high levels of reliability and accessibility across the new systems. Sites will commit to enforce the Compute Canada Security and Privacy Policies at the hosting site, including affected operations personnel. These Policies will include but will not be limited to: physical and logical access control, security screening, operational security management, internal (i.e. Compute Canada) and external audits. System administrator (root) access on the Proposed System(s) may be granted to CC or regional personnel from outside of their institution. This access will be provided on an as- needed, least- privilege basis to qualified and authorized personnel, in order for Compute Canada to implement best practices in systems management and administration. B.4.2 National Systems and Support After consolidation, most researchers will rely on remote hardware resources. Compute Canada will therefore provide a similar look and feel when accessing each system. This national- level support approach will ensure users are able to get connected to the best system, and get all the support they need, regardless of location or language (English or French). Several ongoing initiatives in this area are expected to mature and be deployed with the new infrastructure: Single sign- on: Whether through a Web browser or command line, Compute Canada is working towards a single username and password for all services. This is in cooperation with the Canadian Access Federation (CAF) project. National monitoring: The new systems will be monitored by a new national operations centre, which will give an improved level of monitoring. Critical services will have 24x7 on- call support. This will include a national issue tracking (ticketing) system; making Compute Canada more resilient to failure, and will enable our geographically distributed staff to bring expertise to bear when problems occur. Distributed systems administration: By applying granular privilege separation, appropriately trained staff members will be able to effect changes on remote systems. Activities such as software installations, password resets, and investigations of failed computational jobs will be undertaken by remote staff members in addition to the four sites planned in this stage- 1 proposal. Common software stack, centralized licensing: The four new systems, and subsequent systems, will have similar mechanisms for installing and maintaining software, using modules and other techniques. This will make it easier for users to be portable across systems, and to rapidly become productive on new systems. Highly credentialed staff members: Compute Canada will embark on training to ensure anyone with elevated access, or who needs to provide specific technical support for the new systems, obtains and maintains appropriate credentials. This will include vendor training, third party training, and certifications. Security profiles: The four new systems, sites, and all personnel who have any sort of elevated privileges will be part of the national- level Compute Canada security enclave. Systems and services will be actively monitored, with defense in depth against any type of attack or accident. The newly formed CC Security Council will oversee this. 50

13 Change management: To maintain consistency across systems, and avoid surprises for users or staff members, there will be per- system and national- level configuration change boards (CCBs). The CCBs will provide oversight and consistency with change management. B.4.3 Defunding Existing Systems Compute Canada undertook a cost- benefit analysis to assess which systems should be defunded as a part of the stage- 1 plan. The terminology is defunded instead of decommissioned because the systems belong to the host institutions, which control their ultimate fate. The cost- benefit analysis took into account many factors, starting with the following well- defined measures: Computing power provided by a given system, measured in Tflops; Cost of electricity (including cooling); Cost of maintenance of the system (not including the maintenance of the data center itself). This allows calculating the total cost per Tflops, as shown in the figure below. Total cost per Tflops for all Compute Canada compute servers online during the fall of Green identifies the servers that will remain funded and operational after stage- 1. This analysis determined that most systems commissioned pre were no longer cost- effective. Based on this analysis, and further taking into account the size and configuration of the various clusters as well as the opportunity to conserve some systems as test beds, CC will stop funding 24 systems, and move out of 12 university data centres, in stage- 1. This represents a loss of capacity of 85,000 cores, from approximately 2.0PF to 1.5PF and a loss of 7 PBs of storage. The list of defunded systems includes one of the largest parallel clusters in the current fleet (GPC) and the largest storage site (Silo). This will still leave 17 existing systems (over 100,000 cores) in operation. All existing systems, including those slated for defunding, will remain in operation until the new stage- 1 capacity is available, in order to allow users and data to be seamlessly migrated. 51

14 B.5 Excellence- Based Access B.5.1 Merit- Based Access As documented in Section A, CC has a policy, which grants access to any eligible Canadian researcher, while allocating approximately 80% of available compute resources through a national merit- based review. This review process includes a technical review, eight separate science panels, and a final multi- disciplinary review committee. As competition has grown for a fixed pool of resources, the number of applications submitted to to the Resource Allocation Competition (RAC) each year has grown from 135 in the fall of 2010, to 348 in the fall of In 2013, a FastTrack stream was introduced for researchers who had received strong science reviews the year before and who were requesting to continue their existing allocation. This is attractive to researchers because it reduces the burden required in submission of a new proposal and helps streamline the process for CC staff. 50 projects took advantage of FastTrack when first introduced in However, the growth in the number and diversity of proposals cannot be sustained without additional streamlining and additional staff support. The running of this competition has put a strain on existing CC staff. To address these operational challenges, for the 2014/15 competition, MSI funding allowed CC to hire a consultant with extensive federal granting council experience to review, document and recommend changes to the RAC process. In addition, a permanent science project manager has been hired (September 2014) with significant responsibilities for running the labour- intensive RAC process. The first of the externally recommended changes to the allocation process has already been implemented in the fall 2014 competition with the creation of a separate Research Platforms and Portals (RPP) competition. Researcher feedback indicated that multi- user platforms, which often serve an international community, should not be evaluated against the same criteria as projects serving the needs of individual researchers. For example, while a one- year allocation may be reasonable for an individual project, a platform may instead require a large multi- year storage allocation, which can be accessed by scientists from around the world. CC awarded 13 RPPs in the first competition and expects this competition to grow the list of supported platforms and portals in future years. Another of the key external recommendations was to develop a project plan for the allocation process with detailed timeline and milestones throughout the year. This has been implemented and planning for the fall 2015 competition launch is well underway at the time of writing. Given the rapid growth in allocation applications, it is vital that CC continue to streamline administrative aspects of the process. B.5.2 Support for Contributed Systems In parallel with funding CC, the CFI continues to receive proposals for the funding of advanced computing infrastructure in connection with specific research- focussed projects. In 2012, the CFI modified its Policy and Program Guide to address the housing and managing of any ARC infrastructure to be funded by CFI awards. The so- called Compute Canada Clause indicates a requirement to consult with CC to determine if the infrastructure described in the project can be provided by CC, integrated into CC facilities, or if the infrastructure must or should be separate from CC facilities. A single consultation usually involves a teleconference between project representatives and CC, as well as the exchange of detailed documentation, before the proposal is submitted. It may involve detailed follow- up between project and CC technical teams, discussions with host data centre teams and work on system design. After the award is granted, CC follows- up with all awarded projects that have an identified CC role, for example as an infrastructure host. Since this change of policy, CC has consulted on 91 smaller proposals (CFI LOF/JELF competitions) in which a total of nearly $10M in ARC infrastructure was proposed. In addition, CC consulted with 59 larger projects as part of the recent CFI Innovation Fund (IF) competition. Overall, integration was recommended in 71 out of 52

15 these 150 cases. The awards for the IF competition have only recently been announced, with 13 successful proposals having conditions associated with Compute Canada. The first 47 systems recommended for integration are relatively small, with an incremental power cost for CC estimated at roughly $200,000 in year 4. The IF projects are larger and more complex than JELF awards and the full impact on CC operations is not yet quantified. However, Section C of this document includes a power draw estimate of $600,000 per year for contributed systems starting in MSI year 5 to account for this growth. The cost in staff time for the 150 consultations since 2012 has not been fully estimated. The role of CC in this process has changed throughout the last 2 years (in a positive way) and the systems and procedures to deal with this flow of grants were not in place at the beginning of the MSI period. As CC executes the technology refresh plan described in this document, CC will continue to support pre- and post- proposal consultations with research teams in order to consider, and promote, options for integration. As new systems are brought into service, direct service integration will increasingly be possible. At the same time, CC is seeing an increase in integration discussions, with the recently announced Innovation Fund awards as well as the upcoming Challenge 1 Cyberinfrastructure Initiative. CC will continue to encourage integration of these project- specific facilities into the overall CC infrastructure plan. Research projects benefit from this approach by accessing CC s skilled technical teams, exploiting CC s economies of scale in purchasing, and, where full integration is possible, accessing additional resources when they are needed, rather than being constrained by their own hardware. These research teams receive priority access to a defined amount of ARC resources, and other users may benefit from any spare cycles that may be available from time to time. B.6 Proposed Infrastructure Investment and Options B.6.1 Compute System Types The new systems will be deployed at by mid 2016 at four CC member sites, selected through a national competition. In planning these acquisitions, Compute Canada described two general types of systems. These types distinguish, to some extent, the hardware mix of the systems, and are useful for describing how the systems will be configured and allocated. The two system types are: 1. Large Parallel (LP): a tightly- coupled parallel supercomputer, optimized for running large message passing jobs (i.e., MPI), focused on serving applications using 512 processor cores or more in any single parallel job. This type of system will have a high- speed interconnect and a relatively homogeneous set of computational nodes, with relatively low requirements on memory/node. It will have a tightly coupled high performance parallel file system. It will primarily be used for batch (i.e., non- interactive) jobs. 2. General Purpose (GP): a system type optimized for running a wide range of applications including serial computation, as well as parallel applications spanning a relatively small number of nodes. This type of system may be comprised of a heterogeneous set of nodes (e.g., some with large memory, some with GPUs) and will be well suited to data- intensive applications. They might be suitable for virtualization, Web services, databases, or other tasks that are not primarily compute- intensive. Based on the assessment of needs presented, there is a clear requirement for at least one LP system. This system will have approximately 4 GB of RAM per processor core and a homogeneous configuration. High performance parallel storage suitable for the input/output requirements of the large parallel jobs will be purchased with this system. The aim is to have this system mainly run large parallel jobs. This emphasis on functionality allows for more efficient scheduling of jobs requiring larger numbers of cores than what CC provides today. Users requiring parallel jobs at this scale include those performing computational fluid dynamics calculations (e.g., 53

16 aircraft design, plasma physics, stellar evolution), ocean and atmospheric modelling, and some materials science calculations. General Purpose (GP) systems will serve researchers with a wide range of needs, including those with very large data requirements. These researchers either run serial jobs, jobs that run on single computing nodes or jobs that use a small number of nodes for message passing applications. An increasing number of these jobs also require access to large amounts of data and have high input/output demands. These data- centric jobs have placed considerable demands on the current systems and replacement systems must have suitable input/output performance to address this issue. Many users have a mix of types of jobs that make- up the overall workflow to produce their science output, leading to a preference for systems with a mix of capabilities at a single site. GP systems will be comprised of nodes of different memory sizes. Furthermore, in order to address the changing needs of the CC user base, these systems will also include accelerators (e.g., GPUs) and the capability to support virtualization and containers. Currently the majority of jobs that utilize GPUs run on individual nodes, which is why the current recommendation is that the GPUs be placed in the GP systems. GP systems will also be designed to run virtualized environments and will host some shared storage. The GP systems will also be architected in such a way that at least two security zones are available on each system. These zones will permit isolation of users (and data) with stringent data privacy needs from general- purpose usage. This will allow for some limited support of data with higher security requirements than is required for the majority of CC users. At least one of the GP systems may be designated for even higher security datasets. For researchers requiring high availability for access to their applications and data, the GP systems will be capable of mirroring data across sites, and of having automated failover of applications. B.6.2 Storage Infrastructure As described earlier, needs for storage are acute. Compute Canada s storage as of early 2015 is nearly entirely allocated, and many of the storage subsystems are reaching end of life or end of vendor support. Meanwhile, many of the activities newly funded by CFI have been instructed to look to Compute Canada for their cyberinfrastructure, and storage needs are often at the forefront of requirements. These and other pending needs include very large- scale datasets, notably CANFAR, ATLAS and several genomics projects. Some projects require petabytes of storage and high levels of resiliency and availability. Other storage needs are not as large, but can be costly to implement. For example, several EOIs for the CFI Cyberinfrastructure Challenge 1 Stage- 1 mention data isolation, secure and auditable provisioning of access to data, highly- available data sets, and large databases. These and other characteristics will yield a somewhat different mix of technologies, with different administrative practices than are typical for today s CC operations. (This is discussed further in infrastructure Option 3, below.) For the sites receiving the new compute systems, the high- speed network will be a critical mechanism for data transport and will allow less replication of storage subsystems. In the near term, CC will seek to deploy new storage resources before the associated computing resources. It is hoped that the non- parallel storage (i.e., disk space for block or object access) will be acquired and in place by early This will accommodate some of the storage needs of existing CC users. It will also let CC better serve needs of users and projects that we are in current discussions with. We will use CFI Challenge 1 Competition 1 EOIs to guide some likely futures for storage use. During upcoming months, CC will develop and deploy a storage policy for its resources. A number of policies need to be articulated for the new storage systems and related infrastructure: CC does not provide archival storage, though it can provide long- term storage. 54

17 CC makes reasonable endeavours to protect against data loss or other mishap, but does not guarantee data availability or accuracy unless specific arrangements are made. CC should not be used as the only storage location for valuable data sets. Storage resources are allocated through the same RAC and default allocations process as computational resources. Stored data is subject to removal after a grace period, if an allocation expires and is not renewed. Multiple types or tiers of storage are available, and each type has an allocation process. These include: o High performance parallel storage, used for temporary storage of active computational campaigns. o Dedicated storage for high input/output operations such as databases. This may include flash or solid- state storage devices. o General shared storage: Shared storage pools for persistent data access. Multiple access mechanisms will be provided, including POSIX file/directory access, object access via S3- style RESTful gateways, Globus file transfer, etc. Not all access mechanisms will be applicable for all storage pools. o Hierarchical storage management (HSM) system: Offloading storage from disk to tape, generally for shared storage, although this may include parallel storage. The HSM tape system is not necessarily at the same location as the disk it offloads. o Backups: Much of the general shared storage will be backed up to tape. The backup tape system is not necessarily as the same location as the disk it backs up. o Redundant copies: For any of the above except temporary storage, data replication will be available if needed between particular data pools. Data stored on parallel file systems are not backed up and are subject to automated purging. Quotas and other mechanisms ensure equitable access to storage resources. CC is developing policies and technologies for effective information lifecycle management. Unlike prior generations of batch- mode computational campaigns, in which the produced data may have been viewed as replicable and somewhat transient, we are seeing increasing numbers of projects in which the data sets are the object of interest for science and for operations. CC is ready for these new users and to meet the storage needs of existing users. An accelerated timeline for storage refresh, sooner than the computational systems, will have an immediate benefit to those users. For new users, including national- level partnerships that are being formed today, CC will shift by early 2016 to have mature and well- documented approaches to data- intensive science. This is part of the shift, described earlier, in which CC has been identified as the key provider of national ARC cyberinfrastructure. B.6.3 Infrastructure Options In the following sub- sections, three distinct infrastructure options are presented. These build on the two general system types mentioned above (LP and GP), and presume that four sites will each receive new equipment. It is important to note that final system configurations will vary somewhat from what is described here depending on technology evolution and price changes by the time of the infrastructure RFP response. Site affinity and preferences will be taken into account, as well as indications of future uses or new trends, such as may be indicated by the CFI s Challenge 1 Competition 1 EOIs. As the current focus is on immediate and pressing needs, with more than 20 current systems slated for defunding, there is little scope for highly experimental infrastructure in stage- 1 planning. As such, a baseline option is presented which takes a balanced approach to meeting the needs of the Canadian research community. This is the option recommended by CC. The two alternative options contain many of the same basic required elements that are present in the baseline option but with modifications to emphasize different use- cases. All options involve trade- offs, since the funding identified for stage- 1 cannot meet the projected community need for all LP or GP workloads. In each of the options, four systems are proposed, three classified as general purpose (GP) and one as large parallel (LP) as described further below. Total electrical power for all four systems is estimated at 1.5MW. 55

18 B Option 1: Baseline As discussed in Section 3, different communities of scientists in Canada have specific ARC needs. This baseline option balances the need of the Canadian ARC community for access to compute, storage and accelerators. The relative proportion of each type of infrastructure is chosen based on a combination of current usage statistics and expectations for future evolving need. This is the option that is most in alignment with CC s strategic goals. It provides resources to the broadest range of users across many disciplines, best supporting the entire research community, while building capacity to meet both current and future research needs. This option is intended to maximize the positive effects of research on the lives of Canadians. In this option, a single large parallel (LP) system is proposed. Buying a single system of this type ensures that it is as large as possible within the available budget, to allow the largest jobs to be run. The scale of this system has been chosen such that it will have approximately the same number of cores as the largest of the current CC systems, with the latest technology. LP is also at least the size of the largest system that will be defunded in this round. It is allocated approximately 30% of the overall budget, which matches the fraction of CPU used by jobs of at least 256 cores in The canonical assumption in the LP design is that it will be configured for jobs of at least 512 cores (19% of usage in 2014). There is therefore an assumption of some migration over time from 256 core to at least 512 core jobs given a system which is better configured for that use- case than any current CC system. The interconnect between nodes is assumed to be at least 40Gbps, likely more, with a balanced network and throughput of around 2 Gb/s/core. The LP is scoped to include both substantial fast storage and access to mid- tier storage suitable for a machine of this size. Memory is expected to be 4GB/core. The remainder of the compute and storage budget in this option would be spent on three general purpose (GP) systems at three separate sites. While the LP system is designed to put the maximum number of cores at a single site, this consideration is less important for a GP system. In fact, several large current GP users require at least two geographically separate sites to ensure that large data sets are always accessible. This means that the absolute minimum number of GP systems that can be purchased in stage- 1 is two, or else Compute Canada cannot transition these users from existing systems to the new systems. We propose to buy 3 GP systems in stage- 1, in which two systems (GP2 and GP3) would provide active redundancy for a mix of jobs, including jobs that will benefit from general- purpose graphics processing units (GPUs). The other, GP1, would be somewhat smaller to enable a mix of small parallel and serial jobs and workflows, including via cloud and virtualization services. GP systems are meant to serve a wide range of users in a single site. As such, the node mix is heterogeneous. A number of GP nodes will have large memory (1TB or larger), suitable for large databases and for single- node jobs. Two of the GPs would have large GPU deployments, while the third would have a small deployment of an alternate accelerator technology. GPU nodes would have up to four GPU devices plus up to two CPUs. Each GP hosts a mixture of disk. The configuration of the GP systems will support workloads requiring virtualization. GP1 is expected to include a significant OpenStack deployment to support cloud- based workloads. While the pricing assumes infiniband interconnect at all sites, the OpenStack site might default to 10Gb interconnect depending on IB technology support and price factors at time of purchase. Two tape libraries will be deployed at two sites, to share the load of hierarchical storage management, and backups. This will provide for multiple copies, and geographic diversity, for those uses that require it. In order to support handling of private datasets (e.g., personal health information), GP sites will be configured with multiple security network zones and will implement physical access control policies in the host data centre. 56

19 The GP1 system will be a cluster of nodes that may be configured as needed for diverse purposes. In this baseline option, GP1 has two main duties. First is to handle smaller parallel jobs, including jobs by users who might not yet have applications that can scale 512 or more cores for the LP system. The GP1 system will also be suitable for allocations of nodes to workloads for cloud, database, Web service, science gateways, containers, and other purposes requested by users. While these workloads may also be placed on GP2 and GP3, we anticipate that GP1 will be designed explicitly for easy and flexible mixes of different use cases. Compute Canada is already engaged with cloud deployments and virtualization, which will be a starting point for GP1. All sites, including LP, will host a small set of nodes with GPUs for visualization purposes. All sites will also be upgraded to 100Gb external network connections in stage- 1. This fast inter- site network will allow remote access to shared disk, duplication of object storage repositories, backups, and other sharing and interconnection of resources. At 100Gb/s, hundreds of TB/day may be exchanged. The table below presents an illustrative summary of the four systems. It shows the types of systems expected to be deployed in the baseline option. Potential hosting sites responded to a similar table during the site selection process. It has been updated to be consistent with more recent pricing analysis. Option 1: Baseline System LP GP1 GP2 GP3 CPU cores 30k+ 8k+ 16k+ 16k+ Parallel storage 2PB+ 0.5PB 0.5PB Block and/or object storage 1PB+ 5.5PB+ 8PB+ 8PB+ GPU nodes (e.g., K80) 4 - for vis. 4 - for vis 8 - alt. arch Large memory nodes Estimated CFI Contribution ($million) B Option 2: Larger LP The LP system described in the baseline option will have modern CPUs, larger memory, a faster interconnect, and other improvements as compared to legacy systems. As such, it will benefit CC s larger users, who will transition from systems scheduled for defunding. On a core- for- core basis, LP will mainly replace those systems and not add core hours. On a node- for- node basis, LP will have fewer nodes than the systems it will replace (due to 2X or so cores/node in newer systems, versus legacy systems). CC sees value in having a larger LP system, at the expense of having smaller GP systems. As indicated in Section B.3, this would favour science use- cases involving large system simulations (e.g., earth and ocean, theoretical astrophysics, computational fluid dynamics, some materials science calculations) over those use- cases involving data mining, image processing, etc. (such as genomics, neuroscience, experimental subatomic physics, observational astronomy). For this option, we would work with the LP site to ensure the match funding remains viable. While the needed personnel to manage a larger system would likely not change, the operational costs (particularly power) would increase. Option 2 is to deploy an LP system approximately 1.75X the size of that in the baseline option: over 52,000 cores, approximately 940 kw, at approximately $7M from CFI. The GP2 and GP3 systems would be reduced 57

20 by approximately $.9M each, and the GP1 by $.5M. Disk subsystems would be adjusted to favour LP. Backup and HSM tape systems would remain the same. The advantage of a larger LP is that it would allow CC more capacity for parallel computational jobs. The larger LP would also more easily accommodate a mix of somewhat smaller parallel jobs, rather than placing the higher target of 512 core jobs as in the baseline option and shunting smaller jobs to GP1 and others. The allocations request growth curve supports having a far larger LP system. The drawback of this option is that other workloads - those anticipated for the GP systems - would have fewer available resources. The larger LP would use nearly half of the available stage- 1 budget. At this funding level, the amount of storage available on GP2+GP3 would not be sufficient to migrate some targeted large projects to those new systems. For example, CC has ongoing commitments through its multi- year Research Platforms and Portals competition in 2016 of 8PB of block storage for the ATLAS, CANFAR, CBRAIN, CHIME and T2K projects. These would, by themselves, saturate the budgeted storage on GP2+GP3. CC would therefore defer migration of some of this workload to stage- 2. This likely includes deferring some of the ATLAS Tier- 2 consolidation described above to late Option 2: Larger LP System LP GP1 GP2 GP3 CPU cores 50k+ 8k+ 9.6k+ 9.6k+ Parallel storage 4PB+.5PB.5PB Block and/or object storage 3PB+ 4PB+ 4PB+ 4PB+ GPU nodes (e.g., K80) 4 - for vis. 4 - for vis 8 - alt. arch Large memory nodes Estimated CFI Contribution (million) $7.0 $2.0 $3.2 $2.8 B Option 3: Emphasis on Data Services and Workflows There is a rapidly growing demand for CC to support additional data- intensive activity, versus compute- intensive. Emphasis on data- intensive activity has been a standard CC offering for years, and has been the focus of many recent innovations and enhancements within CC. Data- intensive activities include database services, Web- based access to data, and data analytic capabilities. In addition, many new projects emphasize on- disk (versus near- line tape) access to large datasets this includes digital humanities, genomics/bioinformatics, neuroscience, astronomy, and subatomic physics. CC sees value in taking a larger step towards serving these needs through refactoring of GP2 and GP3. In this scenario, the GP2 and GP3 system specifications would be altered to have a smaller number of GPU nodes (256 for GP2 and 128 for GP3). Those nodes would serve the intended purpose described in the baseline option. The money from those nodes would instead go towards additional data infrastructure, including additional large memory nodes, database infrastructure, data analytics systems, and additional high- availability infrastructure. These shifts in emphasis would make Compute Canada ready for data- focused demand which, to date, has sometimes been served on systems originally designed for LP- type workloads. For this option, while such short- lived jobs would still occur, we would provide more emphasis on capabilities for long- running and resilient services. Large data stores would benefit from Web- based front- 58

21 ends, doing complex queries and data manipulation. Scientists seeking to serve their communities with on- demand data would be able to host on the GP2/GP3 sites (including with failover or redundancy, if needed), knowing that their services would be available. Partner organizations and agencies in Canada would find CC open for business, to meet high- end demands for data isolation, data integrity, access, and availability. For all options, the GP systems will provide cloud- type services, including node provisioning, virtualization, scheduling, and persistent services. They will also provide batch processing. Option 3 would change the balance for GP2 and GP3 from emphasizing batch- mode simulation to emphasizing persistent services and data- intensive activity. In Options 1 and 2, GP2 and GP3 would have significant computational capacity, augmented by GPUs. In Option 3, computational capacity on those systems would be reduced in order to provide larger and more capable systems for other uses. Hardware emphasis would include some large memory systems (over 1TB) suitable for databases and graph analytics. Flexible long- lived Web- based systems with a range of services would be made available, and would run on high- availability clusters and virtualization services. The data subsystems would be augmented with larger quantities of very fast storage (i.e., solid state arrays), pools of highly resilient storage including storage suitable for large databases, and more disk for block/object pools compared to the baseline options. Note that CC is not proposing to engage in data archiving activity, as part of this or other options. The advantage of this option is that it expands the support of big data and disciplines with emerging computing needs, in alignment with CC s strategic goals. The disadvantage is that larger users would be unable to run as effectively on a computing architecture of this design and would have limited availability of resources. Further, the significant reduction in GPU supply would limit CC s ability to support a number of emerging areas including deep learning in the artificial intelligence community. CPU cores and GPU cores would need to be added in stage- 2. Option 3: Emphasis on Data System LP GP1 GP2 GP3 CPU cores 30k+ 10k+ 12k+ 12k+ Parallel storage 2PB+ 0.5PB 0.5PB Block and/or object storage 1PB+ 8PB+ 10PB+ 10PB+ High performance storage 0.3PB+ 0.5PB+ 0.5PB+ GPU nodes (e.g., K80) 4 - for vis. 4 - for vis 8 - alt. arch Large memory nodes Estimated CFI Contribution ($million) B Summary of Infrastructure Options Careful consideration has been made of the need to balance anticipated needs of the research community based on the consultations to date with the opportunities provided with the Cyberinfrastructure funding. CC believes that any of the three options presented are achievable with the budgets and timelines as discussed, at the four named sites. For stage- 1, the procurement will replace aging equipment, with some adjustment to 59

22 reflect contemporary systems and usage (i.e., cloud- style systems management). By stage- 2 CC will begin to address the growing and evolving needs of the community. B.6.4 Site Selection Process and Outcomes CC embarked in earnest on the process of assessing user needs and developing system configurations that would be responsive to those needs in the spring of By summer 2014, CC had completed the preliminary stages of this analysis and established the outlines of CC s Baseline Option 1 (detailed above). This analysis concluded that four new systems would be required under almost any needs- based scenario. Based on this analysis, in autumn 2014, CC began working with Canadian institutions to identify four institutions with high quality data centres willing to make a 5- year financial and operational commitment to hosting a national system. Ideally, the selection process would identify four hosting sites that would maximize both technical quality and reliability as well as financial support and viability. Due to the nature of the CC funding model, the selection process needed to identify proponents who would provide at least 60% matching for the CFI s 40% funding award for both purchase and operating costs of these new systems. In addition, by virtue of this new process, CC had the opportunity to require compliance with minimum national standards on key operational dimensions such as information security, reliability, access and performance. Site Selection Process An RFP- like process was developed by CC management and regional directors, reviewed by the CC Member Council, and approved by the CC Board. The process was launched on October 20, Potential sites were given a baseline scenario very similar to what is outlined in section (above) in order to assess space, power and cooling requirements for proposed systems. Proponents had to meet a number of minimum mandatory requirements (e.g. minimum power capacity, adherence to CC s national security procedures). Proposals would be scored on three technical dimensions: Data Centre Suitability (30% weighting) includes expandability, power efficiency, cooling technologies, available space, etc. Local Staff (20%) while each centre will be supported by the whole CC team, local expertise and experience is vital to a well- run centre. Facility Management Structure (20%) clear points of contact and lines of responsibility should be identified. An additional 30% of the overall proposal score would be awarded based on an assessment of the strength of the financial commitments offered. Nine proposals were submitted by November 27, CC convened an independent International Expert Review Panel to score each proposal using the 70- point scale available for technical considerations. The expert panel submitted a detailed report and clear recommendations to CC. These recommendations went beyond scoring the sites and included best practices for managing a network of sites of this type. Separate from the independent technical scoring, the 30 points for financial commitments was assigned by a team consisting of CC senior management and the four regional directors. Scoring was based on an in- depth evaluation of the strength of the matching funds, including both capital (generally from Provincial sources) and operating (generally from a mix of sources, including in- kind match) expenses. Financial and technical scores were added to produce an overall score. Sensitivity analysis and what if? scenarios were considered, to ensure selections were resilient under conditions where site characteristics may have been over- or under- rated by the expert panel, or financial commitments had been misunderstood. Financial considerations did affect the overall rank of several proposals, but did not alter the final list of four 60

23 sites, which was identical to the list recommended by the expert panel. CC was therefore able to achieve its ideal outcome, namely the selection of four high quality sites that also maximized financial support for the project. Site Selection Results All selected sites met the requirements of the RFP for financial match for capital and operations, using a mix of provincial and institutional funds. University of Toronto: UofT has a history of operating some of CC s largest parallel computers, and brings a strong and experienced team. The data centre is located in Vaughan, and currently houses several large systems. The site benefits from on- site liquid cooling and air handling systems. As part of the site hosting agreement, the site will be augmented with sufficient backup generator power to allow an orderly shutdown of anticipated future storage and other infrastructure, in the event of a power outage. Pending shutdown of legacy systems, UoT has space, power and cooling for well over 4X the needed capacity for the baseline LP system envisioned for stage- 1. University of Toronto will host LP. Simon Fraser University: SFU s Water Tower building was originally a telecoms operational facility, and benefits from extensive physical security and self- contained infrastructure. The site is undergoing an upgrade to free air heat exchangers, for better efficiency for liquid cooling. The data centre has sufficient space and infrastructure (cooling, power, network) to host a variety of possible system configurations, along with associated storage, with space, power and cooling sufficient for more than 5X the needs of the single largest GP system envisioned in stage- 1. SFU staff expertise is particularly strong for meeting high- level security requirements for particular user groups. SFU will host GP2. University of Waterloo: UW s data centre is purpose- built to the Uptime Institute s Tier 3 specification or greater. It has redundancy for critical equipment and other infrastructure. During upcoming months in preparation for system installation, UW will add liquid cooling capability to the data centre. The data centre building, which was recently acquired by UW and is directly adjacent to campus, includes modern offices and conference rooms, and a theatre- style meeting room. The data center has space, power and cooling sufficient for more than 3X the needs of the envisioned GP systems. UW will host GP3. University of Victoria: UVic has recently commissioned the data centre the new CC system will be housed in. It is purpose- built for research computing. Half of the data centre is equipped with raised floors, backup generators, liquid cooling, air handling (including free air heat exchangers), and 100Gb networking. Space, power and cooling in this first half is sufficient for at least 3X the needs of the envisioned GP systems. The other half is an empty shell, ready for installing the next generation of infrastructure when it is needed in the future. UVic staff members have been involved extensively in deployment of CC s new cloud infrastructure, which will be particularly relevant in making configuration choices for the new systems. UVic will host GP1. B.6.5 National Procurement Process The process of procuring four new systems will involve technical and purchasing representatives from all four selected hosting sites, as well as key experts from across CC, forming a national procurement team. This reflects the expectation that stage- 1 systems will serve the entire spectrum of ARC needs across Canada, and is consistent with CC s commitment to broad participation by qualified staff in future operations, even if they do not work at sites where the new systems will be installed. Each stage- 1 host institution will purchase and own a system. However, as part of the site selection process, each site pledged commitment to a national, collaborative procurement process. Based on early discussions with procurement offices at the four institutions, CC expects to utilize a single Request for Proposals (RFP), or a set of coordinated RFPs with shared sections. There will be sections that describe the four system requirements, the four sites, and any procedures specific to each site. A unified RFP will allow vendors to bid multiple systems and to potentially offer discounts or other benefits if they are selected to provide more than one system. 61

24 Bid evaluation will be conducted by the national procurement team, including representation from all four sites. Each site s procurement office will ensure compliance with local and provincial regulations, and must approve the process and outcomes of the decision for that site. Multiple synergies will be achieved through this national procurement process: Bidders will be motivated to consider the broader national context of their offers; common infrastructure will be purchased collectively when possible; and sites will pool their purchases for items that more than one site needs, such as new network switches. During the team s initial development of the RFP, sites are likely to identify better ways to work together, and to benefit from each others best practices and experiences. B.7 Looking Forward Stage- 2 Cyberinfrastructure and Beyond This document focuses on stage- 1 cyberinfrastructure, however CC s infrastructure planning works in a broader context and over a longer time frame. The current CC strategic plan extends to 2019, and the CFI has already announced a second stage Cyberinfrastructure Initiative with a request for a proposal due in April The stage- 2 proposal will also be required to provide for the needs of Challenge 1, domain specific data proposals, to be awarded in spring CC looks forward to working with the CFI towards future investment opportunities beyond stage- 2 to continue to ensure Canadian researchers are well served in the area of advanced research computing. The goal is to work with partners to achieve regular, sustained capital investment in order to avoid both technological obsolescence and resultant negative impact on researchers. While stage- 1 will bring operational improvements and some updated infrastructure, it will not allow CC to catch- up to the projected demand curves shown earlier in this document. With stage- 2 funding, there should be further progress on the demand curve, while achieving even higher operational efficiency through retirement of older systems. CC is engaged in ongoing planning and activity to remain effective as Canada s provider of shared advanced research computing, associated cyberinfrastructure, and support. Efforts towards increased consolidation of systems, with localized support, are a hallmark of national cyberinfrastructure providers in the world today. CC is involved in a number of other initiatives to keep offerings fresh, and to ensure stakeholders have access to the best possible ARC. Growing power requirements, long- term data storage, and the explosion of open access, big data and simulation as essential tools for modern research and innovation are creating challenging financial pressures that directly affect the competitiveness of Canadian research. B.7.1 Consolidation and Expandability Having fewer, larger systems is a hallmark of modern national- scale cyberinfrastructure. Rather than seeking to deploy and maintain significant ARC resources at many member sites, CC s technology refresh activities will continue to concentrate new investments, maximizing the value to Canadian research by leveraging economies of scale, and maximizing utility for those Canadian users by enabling larger jobs, easier scheduling, support for diverse workflows, and minimizing data movement between systems. As detailed above, 22 current systems have been selected for defunding once the stage- 1 systems are in operation. Stage- 2 will see the decommissioning of most other current CC systems and the replacement of the defunded capacity. This new infrastructure will enhance service for Canadian researchers through expansion of systems purchased in stage- 1, deployment of new systems, or through a blend of expansion and new installations. 62

25 In stage- 1, CC will plan the design and procure systems with expansion in mind. The four hosting data centres have available space, power and cooling to expand. The RFP issued for the stage- 1 systems will include future expandability as a key design parameter for vendors. The enhanced security posture of the organization will also be reflected in both the physical and network security built- in to the stage- 1 host facilities. This should allow some of the community needs identified for stage- 2 to be met in a cost- effective way through expansion of stage- 1 systems. In particular, CC expects significant demand for additional storage, cloud and management of secure datasets to result from the CFI s Cyberinfrastructure Challenge 1 competition this year (based on Expressions of Interest received by the CFI). Building flexibility into stage- 1 system designs will allow CC to meet those needs in stage- 2. It will likely be necessary to deploy some new systems in stage- 2, with those systems potentially deployed at new national sites. CC will conduct another national process in 2015 to identify optimal host institutions for new stage- 2 systems. By 2018, CC expects to operate systems in at a maximum of 5 to 10 data centres compared to the 27 centres hosting active systems today. B.7.2 Support of National Projects Compute Canada supports a number of international data- intensive projects in a variety of disciplines. The role of CC in these projects is expanding. As described earlier in this document, the role of CC in serving Canadian astronomy data to the world is expanding through close collaboration with the CANFAR project, the ATLAS Tier- 1 at TRIUMF is expected to migrate to CC by 2018 and ongoing discussions with Canadian Research Data Centre Network may lead to an increased role for CC in serving important social science datasets to researchers. I There are several other discussions ongoing with various Canadian research organizations that have a relatively minor financial impact in stage- 1, but could lead to significant benefits to Canadian researchers in stage- 2 and beyond. The ability to manage long- term projects of this nature is a key advantage of a dedicated national ARC organization and a predictable multi- year capital funding program. For these new endeavours, benefits of CC s extensive expertise and large, distributed staff are evident to the new partners, and provide synergy and leverage with their existing personnel and funding. B Cloud services Cloud- type services have become dominant for enterprise information technology services and offer similar benefits for ARC. CC has recently deployed two experimental instances of a general purpose OpenStack cloud, which have been doing productive work for early users. The GP1 system mentioned above will transition these experimental cloud activities to fully supported production status. CC has also been engaged in discussions with a large Canadian research university to manage their OpenStack cloud together with the CC cloud in a federated manner. CC is also deploying and developing tools to enable researchers to move seamlessly from one cloud instance to another. This CC Cloud Federation is intended to grow into a national cloud service with partner universities across the country. Cloud- type services have definite advantages over more traditional batch- oriented high performance computing (HPC) systems. Cloud services can be more easily customized for the exact operating system and software stack required by particular users. The services are typically deployed on virtualized systems, which allows them to be long- lived, easily portable and replicable, and isolated as needed from other systems and services. The increasing demand for customized, specialized services and system configurations that has already occurred, and is anticipated to grow through Challenge 1, will largely be met by cloud- type services and configurations. B Future Technologies CC is Canada s national provider of ARC, and as such is engaged in ongoing assessment of trends in scientific discovery and analysis that relies on new technological developments. One of the most impactful trends is that HPC is undergoing rapid transitions towards increased power efficiency, better resiliency of software 63

26 and systems, and additional granularity for how computational subsystems integrate and communicate. These shifts are in support of the press towards exascale systems, which are anticipated to become available within 5-7 years. Such systems are necessary for increased accuracy of models, i.e., more physics, higher resolution, and increased temporal resolution that are necessary for areas such as predicting weather systems, modelling environmental impacts, designing complex products and developing new advanced materials. CC staff actively participates in national and international efforts to transition towards next generations of ARC, focusing on development of a national data cloud and a national compute cloud of heterogeneous resources. These efforts, combined with consolidation of sites and distribution of support, will ensure Canada s researchers have the best available mix of technologies and services, well beyond stage- 2. Technology changes for ARC are happening rapidly, and will require scientists and other users to access support to adapt their workflows to benefit from these new technologies. Contemporary trends towards low- power processors and many core, along with accompanying evolution in numerical libraries and methods, create non- trivial needs to adjust existing applications. Data processing techniques and scientific workflows also need to be adjusted. CC tracks the evolution of technology, and delivers training, interpretation and transition support to users. Industry- wide evolution in technologies, and national- level planning, provide the broad framework in which CC operates. Getting benefit from these changes requires CC staff to work individually with researchers, to understand whether and when to adapt their applications and workflows. The pathway to the future requires a mix of activity including technology tracking and deployment, application updates, consideration and implementation of additional science, and appropriate training and support. In Canada, CC is uniquely positioned to guide this transition. 64

27 65

Compute Canada Technology Briefing

Compute Canada Technology Briefing Compute Canada Technology Briefing November 12, 2015 Introduction Compute Canada, in partnership with regional organizations ACENET, Calcul Québec, Compute Ontario and WestGrid, leads the acceleration

More information

Make the Most of Big Data to Drive Innovation Through Reseach

Make the Most of Big Data to Drive Innovation Through Reseach White Paper Make the Most of Big Data to Drive Innovation Through Reseach Bob Burwell, NetApp November 2012 WP-7172 Abstract Monumental data growth is a fact of life in research universities. The ability

More information

Planning a Successful Cloud Strategy Identify existing assets, assess your business needs, and develop a technical and business plan for your cloud

Planning a Successful Cloud Strategy Identify existing assets, assess your business needs, and develop a technical and business plan for your cloud SOLUTION WHITE PAPER Planning a Successful Cloud Strategy Identify existing assets, assess your business needs, and develop a technical and business plan for your cloud Table of Contents Executive Summary

More information

How To Speed Up A Flash Flash Storage System With The Hyperq Memory Router

How To Speed Up A Flash Flash Storage System With The Hyperq Memory Router HyperQ Hybrid Flash Storage Made Easy White Paper Parsec Labs, LLC. 7101 Northland Circle North, Suite 105 Brooklyn Park, MN 55428 USA 1-763-219-8811 www.parseclabs.com [email protected] [email protected]

More information

CYBERINFRASTRUCTURE FRAMEWORK FOR 21 st CENTURY SCIENCE AND ENGINEERING (CIF21)

CYBERINFRASTRUCTURE FRAMEWORK FOR 21 st CENTURY SCIENCE AND ENGINEERING (CIF21) CYBERINFRASTRUCTURE FRAMEWORK FOR 21 st CENTURY SCIENCE AND ENGINEERING (CIF21) Goal Develop and deploy comprehensive, integrated, sustainable, and secure cyberinfrastructure (CI) to accelerate research

More information

The Next Evolution in Storage Virtualization Management

The Next Evolution in Storage Virtualization Management The Next Evolution in Storage Virtualization Management Global Storage Virtualization Simplifies Management, Lowers Operational Costs By Hitachi Data Systems July 2014 Contents Executive Summary... 3 Introduction...

More information

Radware ADC-VX Solution. The Agility of Virtual; The Predictability of Physical

Radware ADC-VX Solution. The Agility of Virtual; The Predictability of Physical Radware ADC-VX Solution The Agility of Virtual; The Predictability of Physical Table of Contents General... 3 Virtualization and consolidation trends in the data centers... 3 How virtualization and consolidation

More information

I D C T E C H N O L O G Y S P O T L I G H T

I D C T E C H N O L O G Y S P O T L I G H T I D C T E C H N O L O G Y S P O T L I G H T U n i fied Cloud Management Increases IT- as- a - S e r vi c e Ag i l i t y November 2013 Adapted from VMware Unifies Cloud Management Portfolio with a Focus

More information

On Demand Satellite Image Processing

On Demand Satellite Image Processing On Demand Satellite Image Processing Next generation technology for processing Terabytes of imagery on the Cloud WHITEPAPER MARCH 2015 Introduction Profound changes are happening with computing hardware

More information

Technology Insight Series

Technology Insight Series Future Proof Your Data Center with NetApp V-Series Russ Fellows January15, 2010 Technology Insight Series Evaluator Group Copyright 2010 Evaluator Group, Inc. and NetApp Inc. All rights reserved Executive

More information

Virtualization s Evolution

Virtualization s Evolution Virtualization s Evolution Expect more from your IT solutions. Virtualization s Evolution In 2009, most Quebec businesses no longer question the relevancy of virtualizing their infrastructure. Rather,

More information

WHITE PAPER. A Practical Guide to Choosing the Right Clouds Option and Storage Service Levels. www.earthlink.com

WHITE PAPER. A Practical Guide to Choosing the Right Clouds Option and Storage Service Levels. www.earthlink.com WHITE PAPER A Practical Guide to Choosing the Right Clouds Option and Storage Service Levels www.earthlink.com 1 Our job in IT is to provide technology frameworks and an operating model to facilitate but

More information

Radware ADC-VX Solution. The Agility of Virtual; The Predictability of Physical

Radware ADC-VX Solution. The Agility of Virtual; The Predictability of Physical Radware ADC-VX Solution The Agility of Virtual; The Predictability of Physical Table of Contents General... 3 Virtualization and consolidation trends in the data centers... 3 How virtualization and consolidation

More information

Building the Business Case for Cloud: Real Ways Private Cloud Can Benefit Your Organization

Building the Business Case for Cloud: Real Ways Private Cloud Can Benefit Your Organization : Real Ways Private Cloud Can Benefit Your Organization In This Paper Leveraging cloud technology can help drive down costs while enabling service-oriented IT. Private and hybrid cloud approaches improve

More information

New Jersey Big Data Alliance

New Jersey Big Data Alliance Rutgers Discovery Informatics Institute (RDI 2 ) New Jersey s Center for Advanced Computation New Jersey Big Data Alliance Manish Parashar Director, Rutgers Discovery Informatics Institute (RDI 2 ) Professor,

More information

ATA DRIVEN GLOBAL VISION CLOUD PLATFORM STRATEG N POWERFUL RELEVANT PERFORMANCE SOLUTION CLO IRTUAL BIG DATA SOLUTION ROI FLEXIBLE DATA DRIVEN V

ATA DRIVEN GLOBAL VISION CLOUD PLATFORM STRATEG N POWERFUL RELEVANT PERFORMANCE SOLUTION CLO IRTUAL BIG DATA SOLUTION ROI FLEXIBLE DATA DRIVEN V ATA DRIVEN GLOBAL VISION CLOUD PLATFORM STRATEG N POWERFUL RELEVANT PERFORMANCE SOLUTION CLO IRTUAL BIG DATA SOLUTION ROI FLEXIBLE DATA DRIVEN V WHITE PAPER Create the Data Center of the Future Accelerate

More information

Optimizing the Data Center for Today s Federal Government

Optimizing the Data Center for Today s Federal Government WHITE PAPER: OPTIMIZING THE DATA CENTER FOR TODAY S FEDERAL......... GOVERNMENT............................... Optimizing the Data Center for Today s Federal Government Who should read this paper CIOs,

More information

CHAPTER - 5 CONCLUSIONS / IMP. FINDINGS

CHAPTER - 5 CONCLUSIONS / IMP. FINDINGS CHAPTER - 5 CONCLUSIONS / IMP. FINDINGS In today's scenario data warehouse plays a crucial role in order to perform important operations. Different indexing techniques has been used and analyzed using

More information

The Methodology Behind the Dell SQL Server Advisor Tool

The Methodology Behind the Dell SQL Server Advisor Tool The Methodology Behind the Dell SQL Server Advisor Tool Database Solutions Engineering By Phani MV Dell Product Group October 2009 Executive Summary The Dell SQL Server Advisor is intended to perform capacity

More information

Simplified Management With Hitachi Command Suite. By Hitachi Data Systems

Simplified Management With Hitachi Command Suite. By Hitachi Data Systems Simplified Management With Hitachi Command Suite By Hitachi Data Systems April 2015 Contents Executive Summary... 2 Introduction... 3 Hitachi Command Suite v8: Key Highlights... 4 Global Storage Virtualization

More information

Building a Scalable Big Data Infrastructure for Dynamic Workflows

Building a Scalable Big Data Infrastructure for Dynamic Workflows Building a Scalable Big Data Infrastructure for Dynamic Workflows INTRODUCTION Organizations of all types and sizes are looking to big data to help them make faster, more intelligent decisions. Many efforts

More information

Cloud Infrastructure Operational Excellence & Reliability

Cloud Infrastructure Operational Excellence & Reliability Cloud Infrastructure Operational Excellence & Reliability Page 1 Operational Excellence & Reliability Microsoft has invested over $15 billion in building one of the world s largest global cloud infrastructures.

More information

A Look at the New Converged Data Center

A Look at the New Converged Data Center Organizations around the world are choosing to move from traditional physical data centers to virtual infrastructure, affecting every layer in the data center stack. This change will not only yield a scalable

More information

Big Data - Infrastructure Considerations

Big Data - Infrastructure Considerations April 2014, HAPPIEST MINDS TECHNOLOGIES Big Data - Infrastructure Considerations Author Anand Veeramani / Deepak Shivamurthy SHARING. MINDFUL. INTEGRITY. LEARNING. EXCELLENCE. SOCIAL RESPONSIBILITY. Copyright

More information

EMC XtremSF: Delivering Next Generation Performance for Oracle Database

EMC XtremSF: Delivering Next Generation Performance for Oracle Database White Paper EMC XtremSF: Delivering Next Generation Performance for Oracle Database Abstract This white paper addresses the challenges currently facing business executives to store and process the growing

More information

Data Centric Computing Revisited

Data Centric Computing Revisited Piyush Chaudhary Technical Computing Solutions Data Centric Computing Revisited SPXXL/SCICOMP Summer 2013 Bottom line: It is a time of Powerful Information Data volume is on the rise Dimensions of data

More information

How To Manage Research Data At Columbia

How To Manage Research Data At Columbia An experience/position paper for the Workshop on Research Data Management Implementations *, March 13-14, 2013, Arlington Rajendra Bose, Ph.D., Manager, CUIT Research Computing Services Amy Nurnberger,

More information

The Business Case for Virtualization Management: A New Approach to Meeting IT Goals By Rich Corley Akorri

The Business Case for Virtualization Management: A New Approach to Meeting IT Goals By Rich Corley Akorri The BusinessCase forvirtualization Management: A New ApproachtoMeetingITGoals ByRichCorley Akorri July2009 The Business Case for Virtualization Management: A New Approach to Meeting IT Goals By Rich Corley

More information

Future Directions in Canadian Research Computing: Complexities of Big Data TORONTO RESEARCH MANAGEMENT SYMPOSIUM, DECEMBER 4, 2014

Future Directions in Canadian Research Computing: Complexities of Big Data TORONTO RESEARCH MANAGEMENT SYMPOSIUM, DECEMBER 4, 2014 Future Directions in Canadian Research Computing: Complexities of Big Data TORONTO RESEARCH MANAGEMENT SYMPOSIUM, DECEMBER 4, 2014 1 Role of ARC* Today * Advanced Research Computing New Paradigms Simulation:

More information

Virtual Desktop Infrastructure Optimization with SysTrack Monitoring Tools and Login VSI Testing Tools

Virtual Desktop Infrastructure Optimization with SysTrack Monitoring Tools and Login VSI Testing Tools A Software White Paper December 2013 Virtual Desktop Infrastructure Optimization with SysTrack Monitoring Tools and Login VSI Testing Tools A Joint White Paper from Login VSI and Software 2 Virtual Desktop

More information

Solution Brief Availability and Recovery Options: Microsoft Exchange Solutions on VMware

Solution Brief Availability and Recovery Options: Microsoft Exchange Solutions on VMware Introduction By leveraging the inherent benefits of a virtualization based platform, a Microsoft Exchange Server 2007 deployment on VMware Infrastructure 3 offers a variety of availability and recovery

More information

Whitepaper. The ABC of Private Clouds. A viable option or another cloud gimmick?

Whitepaper. The ABC of Private Clouds. A viable option or another cloud gimmick? Whitepaper The ABC of Private Clouds A viable option or another cloud gimmick? Although many organizations have adopted the cloud and are reaping the benefits of a cloud computing platform, there are still

More information

Virtualized Security: The Next Generation of Consolidation

Virtualized Security: The Next Generation of Consolidation Virtualization. Consolidation. Simplification. Choice. WHITE PAPER Virtualized Security: The Next Generation of Consolidation Virtualized Security: The Next Generation of Consolidation As we approach the

More information

How To Support High Performance Computing

How To Support High Performance Computing 1 of 12 One Hundred Second Congress of the United States of America AT THE FIRST SESSION Begun and held at the City of Washington on Thursday, the third day of January, one thousand nine hundred and ninety-one

More information

Kent State University s Cloud Strategy

Kent State University s Cloud Strategy Kent State University s Cloud Strategy Table of Contents Item Page 1. From the CIO 3 2. Strategic Direction for Cloud Computing at Kent State 4 3. Cloud Computing at Kent State University 5 4. Methodology

More information

can you effectively plan for the migration and management of systems and applications on Vblock Platforms?

can you effectively plan for the migration and management of systems and applications on Vblock Platforms? SOLUTION BRIEF CA Capacity Management and Reporting Suite for Vblock Platforms can you effectively plan for the migration and management of systems and applications on Vblock Platforms? agility made possible

More information

Realizing the True Potential of Software-Defined Storage

Realizing the True Potential of Software-Defined Storage Realizing the True Potential of Software-Defined Storage Who should read this paper Technology leaders, architects, and application owners who are looking at transforming their organization s storage infrastructure

More information

BUILDING A SCALABLE BIG DATA INFRASTRUCTURE FOR DYNAMIC WORKFLOWS

BUILDING A SCALABLE BIG DATA INFRASTRUCTURE FOR DYNAMIC WORKFLOWS BUILDING A SCALABLE BIG DATA INFRASTRUCTURE FOR DYNAMIC WORKFLOWS ESSENTIALS Executive Summary Big Data is placing new demands on IT infrastructures. The challenge is how to meet growing performance demands

More information

Inside Track Research Note. In association with. Enterprise Storage Architectures. Is it only about scale up or scale out?

Inside Track Research Note. In association with. Enterprise Storage Architectures. Is it only about scale up or scale out? Research Note In association with Enterprise Storage Architectures Is it only about scale up or scale out? August 2015 About this The insights presented in this document are derived from independent research

More information

White paper: Unlocking the potential of load testing to maximise ROI and reduce risk.

White paper: Unlocking the potential of load testing to maximise ROI and reduce risk. White paper: Unlocking the potential of load testing to maximise ROI and reduce risk. Executive Summary Load testing can be used in a range of business scenarios to deliver numerous benefits. At its core,

More information

Cloud Computing and Business Intelligence

Cloud Computing and Business Intelligence Database Systems Journal vol. V, no. 4/2014 49 Cloud Computing and Business Intelligence Alexandru Adrian TOLE Romanian American University, Bucharest, Romania [email protected] The complexity of data

More information

IBM PureFlex System. The infrastructure system with integrated expertise

IBM PureFlex System. The infrastructure system with integrated expertise IBM PureFlex System The infrastructure system with integrated expertise 2 IBM PureFlex System IT is moving to the strategic center of business Over the last 100 years information technology has moved from

More information

Scalability and BMC Remedy Action Request System TECHNICAL WHITE PAPER

Scalability and BMC Remedy Action Request System TECHNICAL WHITE PAPER Scalability and BMC Remedy Action Request System TECHNICAL WHITE PAPER Table of contents INTRODUCTION...1 BMC REMEDY AR SYSTEM ARCHITECTURE...2 BMC REMEDY AR SYSTEM TIER DEFINITIONS...2 > Client Tier...

More information

Evolving Datacenter Architectures

Evolving Datacenter Architectures Technology Insight Paper Evolving Datacenter Architectures HP technologies for Cloud ready IT By Russ Fellows January, 2013 Enabling you to make the best technology decisions Evolving Datacenter Architectures

More information

for Oil & Gas Industry

for Oil & Gas Industry Wipro s Upstream Storage Solution for Oil & Gas Industry 1 www.wipro.com/industryresearch TABLE OF CONTENTS Executive summary 3 Business Appreciation of Upstream Storage Challenges...4 Wipro s Upstream

More information

WHITE PAPER. SAS IT Intelligence. Balancing enterprise strategy, business objectives, IT enablement and costs

WHITE PAPER. SAS IT Intelligence. Balancing enterprise strategy, business objectives, IT enablement and costs WHITE PAPER SAS IT Intelligence Balancing enterprise strategy, business objectives, IT enablement and costs Table of Contents Executive summary... 1 SAS IT Intelligence leaping tactical pitfalls... 2 Resource

More information

Reducing Storage TCO With Private Cloud Storage

Reducing Storage TCO With Private Cloud Storage Prepared by: Colm Keegan, Senior Analyst Prepared: October 2014 With the burgeoning growth of data, many legacy storage systems simply struggle to keep the total cost of ownership (TCO) in check. This

More information

Best Practices for Implementing Global IoT Initiatives Key Considerations for Launching a Connected Devices Service

Best Practices for Implementing Global IoT Initiatives Key Considerations for Launching a Connected Devices Service Best Practices for Implementing Global IoT Initiatives Key Considerations for Launching a Connected Devices Service White Paper Jasper Technologies, Inc. 02 For more information about becoming a service

More information

Achieve Economic Synergies by Managing Your Human Capital In The Cloud

Achieve Economic Synergies by Managing Your Human Capital In The Cloud Achieve Economic Synergies by Managing Your Human Capital In The Cloud By Orblogic, March 12, 2014 KEY POINTS TO CONSIDER C LOUD S OLUTIONS A RE P RACTICAL AND E ASY TO I MPLEMENT Time to market and rapid

More information

User-Centric Client Management with System Center 2012 Configuration Manager in Microsoft IT

User-Centric Client Management with System Center 2012 Configuration Manager in Microsoft IT Situation Microsoft IT needed to evolve their Configuration Manager 2007-based environment that used homegrown application distribution services to meet the self-service needs of Microsoft personnel. Solution

More information

ORACLE DATABASE 10G ENTERPRISE EDITION

ORACLE DATABASE 10G ENTERPRISE EDITION ORACLE DATABASE 10G ENTERPRISE EDITION OVERVIEW Oracle Database 10g Enterprise Edition is ideal for enterprises that ENTERPRISE EDITION For enterprises of any size For databases up to 8 Exabytes in size.

More information

Big Data Challenges in Bioinformatics

Big Data Challenges in Bioinformatics Big Data Challenges in Bioinformatics BARCELONA SUPERCOMPUTING CENTER COMPUTER SCIENCE DEPARTMENT Autonomic Systems and ebusiness Pla?orms Jordi Torres [email protected] Talk outline! We talk about Petabyte?

More information

How To Build A Supermicro Computer With A 32 Core Power Core (Powerpc) And A 32-Core (Powerpc) (Powerpowerpter) (I386) (Amd) (Microcore) (Supermicro) (

How To Build A Supermicro Computer With A 32 Core Power Core (Powerpc) And A 32-Core (Powerpc) (Powerpowerpter) (I386) (Amd) (Microcore) (Supermicro) ( TECHNICAL GUIDELINES FOR APPLICANTS TO PRACE 7 th CALL (Tier-0) Contributing sites and the corresponding computer systems for this call are: GCS@Jülich, Germany IBM Blue Gene/Q GENCI@CEA, France Bull Bullx

More information

Brochure. Update your Windows. HP Technology Services for Microsoft Windows 2003 End of Support (EOS) and Microsoft Migrations

Brochure. Update your Windows. HP Technology Services for Microsoft Windows 2003 End of Support (EOS) and Microsoft Migrations Brochure Update your Windows HP Technology Services for Microsoft End of Support (EOS) and Microsoft Migrations Stabilize and secure your infrastructure Microsoft will end support for Windows Server 2003/R2

More information

The Shortcut Guide to Balancing Storage Costs and Performance with Hybrid Storage

The Shortcut Guide to Balancing Storage Costs and Performance with Hybrid Storage The Shortcut Guide to Balancing Storage Costs and Performance with Hybrid Storage sponsored by Dan Sullivan Chapter 1: Advantages of Hybrid Storage... 1 Overview of Flash Deployment in Hybrid Storage Systems...

More information

Solution brief. HP solutions for IT service management. Integration, automation, and the power of self-service IT

Solution brief. HP solutions for IT service management. Integration, automation, and the power of self-service IT Solution brief HP solutions for IT service management Integration, automation, and the power of self-service IT Make IT indispensable to the business. Turn IT staff into efficient, cost-cutting rock stars.

More information

Cisco Network Optimization Service

Cisco Network Optimization Service Service Data Sheet Cisco Network Optimization Service Optimize your network for borderless business evolution and innovation using Cisco expertise and leading practices. New Expanded Smart Analytics Offerings

More information

Making a Case for Including WAN Optimization in your Global SharePoint Deployment

Making a Case for Including WAN Optimization in your Global SharePoint Deployment Making a Case for Including WAN Optimization in your Global SharePoint Deployment Written by: Mauro Cardarelli Mauro Cardarelli is co-author of "Essential SharePoint 2007 -Delivering High Impact Collaboration"

More information

IBM Deep Computing Visualization Offering

IBM Deep Computing Visualization Offering P - 271 IBM Deep Computing Visualization Offering Parijat Sharma, Infrastructure Solution Architect, IBM India Pvt Ltd. email: [email protected] Summary Deep Computing Visualization in Oil & Gas

More information

WHITE PAPER. www.fusionstorm.com. Building Blocks of the Modern Data Center

WHITE PAPER. www.fusionstorm.com. Building Blocks of the Modern Data Center WHITE PAPER: Easing the Way to the Cloud: 1 WHITE PAPER Building Blocks of the Modern Data Center How Integrated Infrastructure Solutions Help to Accelerate Application Deployments and Simplify Management

More information

Azure Scalability Prescriptive Architecture using the Enzo Multitenant Framework

Azure Scalability Prescriptive Architecture using the Enzo Multitenant Framework Azure Scalability Prescriptive Architecture using the Enzo Multitenant Framework Many corporations and Independent Software Vendors considering cloud computing adoption face a similar challenge: how should

More information

How To Save Money On A Data Center

How To Save Money On A Data Center Federal Data Center Consolidation Initiative Data Center Consolidation Plan for the U.S. Small Business Administration Maintained by: The Office of the Chief Information Officer Paul Christy, CIO Revised:

More information

EMC XtremSF: Delivering Next Generation Storage Performance for SQL Server

EMC XtremSF: Delivering Next Generation Storage Performance for SQL Server White Paper EMC XtremSF: Delivering Next Generation Storage Performance for SQL Server Abstract This white paper addresses the challenges currently facing business executives to store and process the growing

More information

The Benefits of Virtualizing

The Benefits of Virtualizing T E C H N I C A L B R I E F The Benefits of Virtualizing Aciduisismodo Microsoft SQL Dolore Server Eolore in Dionseq Hitachi Storage Uatummy Environments Odolorem Vel Leveraging Microsoft Hyper-V By Heidi

More information

Big Workflow: More than Just Intelligent Workload Management for Big Data

Big Workflow: More than Just Intelligent Workload Management for Big Data Big Workflow: More than Just Intelligent Workload Management for Big Data Michael Feldman White Paper February 2014 EXECUTIVE SUMMARY Big data applications represent a fast-growing category of high-value

More information

Data Deduplication: An Essential Component of your Data Protection Strategy

Data Deduplication: An Essential Component of your Data Protection Strategy WHITE PAPER: THE EVOLUTION OF DATA DEDUPLICATION Data Deduplication: An Essential Component of your Data Protection Strategy JULY 2010 Andy Brewerton CA TECHNOLOGIES RECOVERY MANAGEMENT AND DATA MODELLING

More information

Role of Analytics in Infrastructure Management

Role of Analytics in Infrastructure Management Role of Analytics in Infrastructure Management Contents Overview...3 Consolidation versus Rationalization...5 Charting a Course for Gaining an Understanding...6 Visibility into Your Storage Infrastructure...7

More information

EMC PERSPECTIVE. The Private Cloud for Healthcare Enables Coordinated Patient Care

EMC PERSPECTIVE. The Private Cloud for Healthcare Enables Coordinated Patient Care EMC PERSPECTIVE The Private Cloud for Healthcare Enables Coordinated Patient Care Table of Contents A paradigm shift for Healthcare IT...................................................... 3 Cloud computing

More information

Upgrading to Microsoft SQL Server 2008 R2 from Microsoft SQL Server 2008, SQL Server 2005, and SQL Server 2000

Upgrading to Microsoft SQL Server 2008 R2 from Microsoft SQL Server 2008, SQL Server 2005, and SQL Server 2000 Upgrading to Microsoft SQL Server 2008 R2 from Microsoft SQL Server 2008, SQL Server 2005, and SQL Server 2000 Your Data, Any Place, Any Time Executive Summary: More than ever, organizations rely on data

More information

Big Data 101: Harvest Real Value & Avoid Hollow Hype

Big Data 101: Harvest Real Value & Avoid Hollow Hype Big Data 101: Harvest Real Value & Avoid Hollow Hype 2 Executive Summary Odds are you are hearing the growing hype around the potential for big data to revolutionize our ability to assimilate and act on

More information

BACKUP IS DEAD: Introducing the Data Protection Lifecycle, a new paradigm for data protection and recovery WHITE PAPER

BACKUP IS DEAD: Introducing the Data Protection Lifecycle, a new paradigm for data protection and recovery WHITE PAPER BACKUP IS DEAD: Introducing the Data Protection Lifecycle, a new paradigm for data protection and recovery Despite decades of research and development into backup and data protection, enterprise customers

More information

Cluster, Grid, Cloud Concepts

Cluster, Grid, Cloud Concepts Cluster, Grid, Cloud Concepts Kalaiselvan.K Contents Section 1: Cluster Section 2: Grid Section 3: Cloud Cluster An Overview Need for a Cluster Cluster categorizations A computer cluster is a group of

More information

Impact of Big Data in Oil & Gas Industry. Pranaya Sangvai Reliance Industries Limited 04 Feb 15, DEJ, Mumbai, India.

Impact of Big Data in Oil & Gas Industry. Pranaya Sangvai Reliance Industries Limited 04 Feb 15, DEJ, Mumbai, India. Impact of Big Data in Oil & Gas Industry Pranaya Sangvai Reliance Industries Limited 04 Feb 15, DEJ, Mumbai, India. New Age Information 2.92 billions Internet Users in 2014 Twitter processes 7 terabytes

More information

The Massachusetts Open Cloud (MOC)

The Massachusetts Open Cloud (MOC) The Massachusetts Open Cloud (MOC) October 11, 2012 Abstract The Massachusetts open cloud is a new non-profit open public cloud that will be hosted (primarily) at the MGHPCC data center. Its mission is

More information

Bringing Big Data Modelling into the Hands of Domain Experts

Bringing Big Data Modelling into the Hands of Domain Experts Bringing Big Data Modelling into the Hands of Domain Experts David Willingham Senior Application Engineer MathWorks [email protected] 2015 The MathWorks, Inc. 1 Data is the sword of the

More information

Relational Databases in the Cloud

Relational Databases in the Cloud Contact Information: February 2011 zimory scale White Paper Relational Databases in the Cloud Target audience CIO/CTOs/Architects with medium to large IT installations looking to reduce IT costs by creating

More information

CITY OF VAUGHAN EXTRACT FROM COUNCIL MEETING MINUTES OF FEBRUARY 17, 2015

CITY OF VAUGHAN EXTRACT FROM COUNCIL MEETING MINUTES OF FEBRUARY 17, 2015 EXTRACT FROM COUNCIL MEETING MINUTES OF FEBRUARY 17, 2015 Item 3, Report No. 5, of the Finance, Administration and Audit Committee, which was adopted without amendment by the Council of the City of Vaughan

More information

A High Performance Computing Scheduling and Resource Management Primer

A High Performance Computing Scheduling and Resource Management Primer LLNL-TR-652476 A High Performance Computing Scheduling and Resource Management Primer D. H. Ahn, J. E. Garlick, M. A. Grondona, D. A. Lipari, R. R. Springmeyer March 31, 2014 Disclaimer This document was

More information

Joe Young, Senior Windows Administrator, Hostway

Joe Young, Senior Windows Administrator, Hostway Many of our enterprise customers wanted dedicated virtual servers that offered a much higher degree of isolation... we needed to step up our virtualization efforts to stay competitive." Joe Young, Senior

More information