Enabling the Use of Data



Similar documents
Achieving Real-Time Business Solutions Using Graph Database Technology and High Performance Networks

Data center day. a silicon photonics update. Alexis Björlin. Vice President, General Manager Silicon Photonics Solutions Group August 27, 2015

Enabling High performance Big Data platform with RDMA

Solving I/O Bottlenecks to Enable Superior Cloud Efficiency

Storage, Cloud, Web 2.0, Big Data Driving Growth

Michael Kagan.

Unstructured Data Accelerator (UDA) Author: Motti Beck, Mellanox Technologies Date: March 27, 2012

Comparing SMB Direct 3.0 performance over RoCE, InfiniBand and Ethernet. September 2014

Mellanox Academy Online Training (E-learning)

SMB Direct for SQL Server and Private Cloud

Simplifying Big Data Deployments in Cloud Environments with Mellanox Interconnects and QualiSystems Orchestration Solutions

Driving IBM BigInsights Performance Over GPFS Using InfiniBand+RDMA

Building a Scalable Storage with InfiniBand

Choosing the Best Network Interface Card for Cloud Mellanox ConnectX -3 Pro EN vs. Intel XL710

Introduction to Cloud Design Four Design Principals For IaaS

LS DYNA Performance Benchmarks and Profiling. January 2009

RoCE vs. iwarp Competitive Analysis

Hadoop: Embracing future hardware

LS-DYNA Best-Practices: Networking, MPI and Parallel File System Effect on LS-DYNA Performance

Can High-Performance Interconnects Benefit Memcached and Hadoop?

I/O Virtualization Using Mellanox InfiniBand And Channel I/O Virtualization (CIOV) Technology

Data Centric Systems (DCS)

SX1024: The Ideal Multi-Purpose Top-of-Rack Switch

Achieving Mainframe-Class Performance on Intel Servers Using InfiniBand Building Blocks. An Oracle White Paper April 2003

Towards Rack-scale Computing Challenges and Opportunities

Advancing Applications Performance With InfiniBand

Hadoop on the Gordon Data Intensive Cluster

Mellanox Accelerated Storage Solutions

Next Generation Operating Systems

Building a Scalable Big Data Infrastructure for Dynamic Workflows

Big Data Performance Growth on the Rise

HPC & Big Data THE TIME HAS COME FOR A SCALABLE FRAMEWORK

Advanced Core Operating System (ACOS): Experience the Performance

White Paper Solarflare High-Performance Computing (HPC) Applications

Open Ethernet. April

ECLIPSE Best Practices Performance, Productivity, Efficiency. March 2009

Windows TCP Chimney: Network Protocol Offload for Optimal Application Scalability and Manageability

SMB Advanced Networking for Fault Tolerance and Performance. Jose Barreto Principal Program Managers Microsoft Corporation

Big Data Management in the Clouds and HPC Systems

ECLIPSE Performance Benchmarks and Profiling. January 2009

Netvisor Software Defined Fabric Architecture

Data Center Network Evolution: Increase the Value of IT in Your Organization

Architecting Low Latency Cloud Networks

An Oracle White Paper June High Performance Connectors for Load and Access of Data from Hadoop to Oracle Database

Interconnecting Future DoE leadership systems

Optimizing Web Infrastructure on Intel Architecture

Comprehensive Analytics on the Hortonworks Data Platform

Next Steps Toward 10 Gigabit Ethernet Top-of-Rack Networking

Switching Architectures for Cloud Network Designs

VirtualclientTechnology 2011 July

Block based, file-based, combination. Component based, solution based

Realizing the next step in storage/converged architectures

AlcAtel-lucent enterprise AnD sdnsquare sdn² network solution enabling highly efficient, volumetric, time-critical data transfer over ip networks

Deploying Ceph with High Performance Networks, Architectures and benchmarks for Block Storage Solutions

Solving the Hypervisor Network I/O Bottleneck Solarflare Virtualization Acceleration

High Performance OpenStack Cloud. Eli Karpilovski Cloud Advisory Council Chairman

Data Centric Computing Revisited

Redefining the Data Center Edge

ECMWF HPC Workshop: Accelerating Data Management

OPTIMIZING SERVER VIRTUALIZATION

Storage at a Distance; Using RoCE as a WAN Transport

Jean-Pierre Panziera Teratec 2011

Vendor Update Intel 49 th IDC HPC User Forum. Mike Lafferty HPC Marketing Intel Americas Corp.

Unified Computing Systems

SX1012: High Performance Small Scale Top-of-Rack Switch

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

Connecting the Clouds

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

Cloud Data Center Acceleration 2015

InfiniBand vs Fibre Channel Throughput. Figure 1. InfiniBand vs 2Gb/s Fibre Channel Single Port Storage Throughput to Disk Media

INFINIBAND S DATA CENTER MARCH

Solid State Storage in the Evolution of the Data Center

Modernizing Hadoop Architecture for Superior Scalability, Efficiency & Productive Throughput. ddn.com

New Features in SANsymphony -V10 Storage Virtualization Software

Hur hanterar vi utmaningar inom området - Big Data. Jan Östling Enterprise Technologies Intel Corporation, NER

Choosing the Best Network Interface Card Mellanox ConnectX -3 Pro EN vs. Intel X520

Parallel Computing: Strategies and Implications. Dori Exterman CTO IncrediBuild.

IBM Netezza High Capacity Appliance

Scaling Cloud-Native Virtualized Network Services with Flash Memory

State of the Art Cloud Infrastructure

Switch Chip panel discussion. Moderator: Yoshihiro Nakajima (NTT)

The Future of Cloud Networking. Idris T. Vasi

The Software Defined Hybrid Packet Optical Datacenter Network SDN AT LIGHT SPEED TM CALIENT Technologies

Installing Hadoop over Ceph, Using High Performance Networking

How To Build A Cloud Computer

Microsoft Windows Server Hyper-V in a Flash

BUILDING A SCALABLE BIG DATA INFRASTRUCTURE FOR DYNAMIC WORKFLOWS

What s New in Mike Bailey LabVIEW Technical Evangelist. uk.ni.com

Big Data Challenges in Bioinformatics

Intel Data Direct I/O Technology (Intel DDIO): A Primer >

Oracle s Big Data solutions. Roger Wullschleger. <Insert Picture Here>

Replacing SAN with High Performance Windows Share over a Converged Network

Can Flash help you ride the Big Data Wave? Steve Fingerhut Vice President, Marketing Enterprise Storage Solutions Corporation

Transcription:

Enabling the Use of Data Michael Kagan, CTO June 1, 2015 - Technion Computer Engineering Conference

Safe Harbor Statement These slides and the accompanying oral presentation contain forward-looking statements and information. The use of words such as may, might, will, should, expect, plan, anticipate, believe, estimate, project, intend, future, potential or continued, and other similar expressions are intended to identify forward-looking statements. All of these forward-looking statements are based on estimates and assumptions by our management that, although we believe to be reasonable, are inherently uncertain. Forward-looking statements involve risks and uncertainties, including, but not limited to, economic, competitive, governmental and technological factors outside of our control, that may cause our business, industry, strategy or actual results to differ materially from the forward-looking statement. These risks and uncertainties may include those discussed under the heading Risk Factors in the Company s most recent 10K and 10Qs on file with the Securities and Exchange Commission, and other factors which may not be known to us. Any forward-looking statement speaks only as of its date. We undertake no obligation to publicly update or revise any forward-looking statement, whether as a result of new information, future events or otherwise, except as required by law. 2015 Mellanox Technologies 2

Data Explosion 2015 Mellanox Technologies 3

Data Processing Plant Convert Data to Information 2015 Mellanox Technologies 4

Create System Out of Compute & Storage Networks APPLICATIONS NETWORKS COMPUTING, STORAGE, APPLIANCES The Future Depends on Mellanox 2015 Mellanox Technologies 5

Mellanox Enables Scale, Robustness, Performance Data-Intensive Simulations Internet of Things Mission Critical National Security Healthcare Machine Learning Smart Cities Business Intelligence 2015 Mellanox Technologies 6

Data Center Evolution Over Time Multi-Server Multi-Core Multi-Host 2015 Mellanox Technologies 7

Data Center Evolution Data center North-south traffic 80% Web Server Data center Data center East-west traffic 70% App Server Database Storage 2015 Mellanox Technologies 8

Compute-Centric Data Center Compute Centric Center Architecture Networking Data I/O?? Memory Resources Memory Memory CPU Compute Storage Building Block Compute Node 2015 Mellanox Technologies 9

Data-Centric Data Center Data-Centric Data Center Architecture Networking Compute Storage Data Building Block Cluster 2015 Mellanox Technologies 10

Rack & Data Center Level Optimization for the Cloud Rack Level Optimization For The Data Centric Data Center Multi-Host Adapter High Speed Data Connectivity NIC Server Disaggregation Silicon Photonics 2015 Mellanox Technologies 11

Efficient Data Movement CPU Onload Network Offload x x x x CPU Onload Penalties Half the Throughput Twice the Latency 2X CPU Consumption Efficient Data Movement With RDMA 2X Better Bandwidth Half the Latency 2X Better CPU Efficiency 2015 Mellanox Technologies 12

Mellanox Multi-Host Technology Ideal for ARM CPUs CPU CPU CPU Lower Your Connectivity Cost by 45%! CPU CPU CPU CPU CPU Traditional Design Multi-Host Technology 2015 Mellanox Technologies 13

Multi-Host Dramatically Reduces Server Cost Traditional 4-Way SMP Machine Multi-Host 4-Socket Architecture CPU CPU CPU CPU CPU CPU CPU CPU Modern CPUs with 8-20 cores don t require expensive SMP architectures. Additional parallelism achieved with higher level network based distributed programming techniques such as Hadoop Map-Reduce 2015 Mellanox Technologies 14

ConnectX-4 on Facebook OCP Multi-Host Platform (Yosemite) ConnectX-4 Multi-Host Adapter Compute Slots OCP is Ideal Open Multi-Host Platform for ARM CPUs The Next Generation Compute and Storage Rack Design 2015 Mellanox Technologies 15

The Future is Here Enter the World of Scalable Performance At the Speed of 100Gb/s! 2015 Mellanox Technologies 16

The Future is Here Entering the Era of 100Gb/s 100Gb/s Adapter, 0.7us latency 150 million messages per second (10 / 25 / 40 / 50 / 56 / 100Gb/s) 36 EDR (100Gb/s) Ports, <90ns Latency Throughput of 7.2Tb/s Copper (Passive, Active) Optical Cables (VCSEL) Silicon Photonics 2015 Mellanox Technologies 17

Enabling Tomorrow s Platforms Today Enabling the Use of Data 2015 Mellanox Technologies 18

Thank You

Exponential Data Growth Requires a Faster Cloud Data Centers We Live in a World of Data More Devices More Applications More Data Data Needs to be Accessible Always and in Real-Time 2015 Mellanox Technologies 20