The Performance of 2 and 4 HyperDrive4 units in RAID0



Similar documents
Education. Servicing the IBM ServeRAID-8k Serial- Attached SCSI Controller. Study guide. XW5078 Release 1.00

HP Z Turbo Drive PCIe SSD

Adaptec: Snap Server NAS Performance Study

Intel Solid- State Drive Data Center P3700 Series NVMe Hybrid Storage Performance

Communicating with devices

Chapter Introduction. Storage and Other I/O Topics. p. 570( 頁 585) Fig I/O devices can be characterized by. I/O bus connections

Tekla Structures 18 Hardware Recommendation

F600Q 8Gb FC Storage Performance Report Date: 2012/10/30

Optimizing Computer Systems To Run SolidWorks. Jeff Setzer SolidWorks / 3DVIA Product Manager Graphics Systems Corporation

Performance Report Modular RAID for PRIMERGY

AP ENPS ANYWHERE. Hardware and software requirements

How to choose a suitable computer

PCI SATA RAID Card User Manual

Memoright SSDs: The End of Hard Drives?

Adaptec: Snap Server NAS Performance Study

Accelerating Server Storage Performance on Lenovo ThinkServer

Several tips on how to choose a suitable computer

Performance Characteristics of VMFS and RDM VMware ESX Server 3.0.1

Quantifying Hardware Selection in an EnCase v7 Environment

Computer Performance. Topic 3. Contents. Prerequisite knowledge Before studying this topic you should be able to:

SSDs and RAID: What s the right strategy. Paul Goodwin VP Product Development Avant Technology

Enterprise Edition. Hardware Requirements

SSD Old System vs HDD New

Price/performance Modern Memory Hierarchy

O&O Defrag and the Windows Defragmenter: A comparison

Several tips on how to choose a suitable computer

OBJECTIVE ANALYSIS WHITE PAPER MATCH FLASH. TO THE PROCESSOR Why Multithreading Requires Parallelized Flash ATCHING

The Bus (PCI and PCI-Express)

Embedded Operating Systems in a Point of Sale Environment. White Paper

Intel Solid-State Drives Increase Productivity of Product Design and Simulation

PCIE High performance dual core Pentium D with SATAII Raid 0,1,5,10 function support!

Profit Tools for Trucking General System Requirements. Requirements For Hardware, Network, Software and Communications

Distribution One Server Requirements

STI Hardware Specifications for PCs

Sistemas Operativos: Input/Output Disks

Virtuoso and Database Scalability

Models Smart Array 6402A/128 Controller 3X-KZPEC-BF Smart Array 6404A/256 two 2 channel Controllers

Fusionstor NAS Enterprise Server and Microsoft Windows Storage Server 2003 competitive performance comparison

User manual Ver: 2.0. SATA Channel PCI RAID Card

Understanding the Performance of an X User Environment

Priority Pro v17: Hardware and Supporting Systems

Dynamode External USB3.0 Dual RAID Encloure. User Manual.

Executive Summary: Test methodology:

Grant Management. System Requirements

DIABLO TECHNOLOGIES MEMORY CHANNEL STORAGE AND VMWARE VIRTUAL SAN : VDI ACCELERATION

Qsan Document - White Paper. Performance Monitor Case Studies

White paper. QNAP Turbo NAS with SSD Cache

Maximizing VMware ESX Performance Through Defragmentation of Guest Systems. Presented by

RAID. RAID 0 No redundancy ( AID?) Just stripe data over multiple disks But it does improve performance. Chapter 6 Storage and Other I/O Topics 29

Architecting High-Speed Data Streaming Systems. Sujit Basu

Certification Document bluechip STORAGEline R54300s NAS-Server 03/06/2014. bluechip STORAGEline R54300s NAS-Server system

Introduction to I/O and Disk Management

SSDs tend to be more rugged than hard drives with respect to shock and vibration because SSDs have no moving parts.

Dell Microsoft Business Intelligence and Data Warehousing Reference Configuration Performance Results Phase III

H ARDWARE C ONSIDERATIONS

Benefits of Intel Matrix Storage Technology

RAMDISK Benchmarks Test system: Software OS Setup Hypothesis License Author

Hard Disk Drive vs. Kingston SSDNow V+ 200 Series 240GB: Comparative Test

Choosing a Computer for Running SLX, P3D, and P5

Simnet Registry Repair User Guide. Edition 1.3

Agenda. Enterprise Application Performance Factors. Current form of Enterprise Applications. Factors to Application Performance.

QuickSpecs. HP Smart Array 5312 Controller. Overview

Q & A From Hitachi Data Systems WebTech Presentation:

SiS AMD Athlon TM 64FX/PCI-E Solution. Silicon Integrated Systems Corp. Integrated Product Division April. 2004

Serial ATA RAID PCI. User's Manual

Sun Microsystems Special Promotions for Education and Research January 9, 2007

Qualified Apple Mac Workstations for Avid Media Composer v5.0.x

Tested product: Auslogics BoostSpeed

RAID HARDWARE. On board SATA RAID controller. RAID drive caddy (hot swappable) SATA RAID controller card. Anne Watson 1

Comparison of best notebooks

HP reference configuration for entry-level SAS Grid Manager solutions

Abila Grant Management. System Requirements

PCI vs. PCI Express vs. AGP

File System & Device Drive. Overview of Mass Storage Structure. Moving head Disk Mechanism. HDD Pictures 11/13/2014. CS341: Operating System

SATA RAID Function (Only for chipset Sil3132 used) User s Manual

Desktop PC Buying Guide

HP Smart Array Controllers and basic RAID performance factors

Maximizing SQL Server Virtualization Performance

RAID 5 rebuild performance in ProLiant

UK HQ RAID Chunk Size T F ISO 14001

SSDs tend to be more rugged than hard drives with respect to shock and vibration because SSDs have no moving parts.

Terminal Server Software and Hardware Requirements. Terminal Server. Software and Hardware Requirements. Datacolor Match Pigment Datacolor Tools

Certification Document macle GmbH GRAFENTHAL R2208 S2 01/04/2016. macle GmbH GRAFENTHAL R2208 S2 Storage system

Workstation Virtualization Software Review. Matthew Smith. Office of Science, Faculty and Student Team (FaST) Big Bend Community College

Autodesk Building Design Suite 2012 Standard Edition System Requirements... 2

HP Proliant BL460c G7

PMC Solid State Memory Card Product Manual V

SATA II 4 Port PCI RAID Card RC217 User Manual

Intel RAID Software User s Guide:

WHITE PAPER FUJITSU PRIMERGY SERVER BASICS OF DISK I/O PERFORMANCE

H ARDWARE C ONSIDERATIONS

Data Sheet ESPRIMO P1500. The wide-ranging home office and multimedia PC ESPRIMO P1500

Business white paper. HP Process Automation. Version 7.0. Server performance

s y s t e m r e q u i r e m e n t s

MedInformatix System Requirements

Transcription:

The Performance of 2 and 4 HyperDrive4 units in RAID0 The DDR in a HyperDrive4 is only run at 100 MHz in order to save energy and prolong life and reliability. Even at this sedate pace, since it has a 64 bit bus (well 72 with the parity check, but 64 form a performance standpoint) it is effectively PC800. So it is capable of ATA800 performance. The HyperDrive does not achieve this performance of course since it has an ATA133 bus. But despite having an industry standard bus designed for mechanical hard disks, due to the fact that the HyperDrive4 has a seek time of 1.1 microseconds read and 250 nanoseconds write, the device is twice as fast as any number of WD740ADFD Raptors in RAID6 aged for defragmentation. But we want to get closer to the full performance potential of DDR storage in order that all the time in our lives that is wasted by parasitic public sector bureacrats can be saved by expensive private sector technology. So the obvious thing to do is try 2 and 4 HyperDrives in RAID0. One has to be careful when choosing a RAID card for 4 HyperDrives. Firstly since the STR of each unit is 120 MB per second, the total theoretical STR is 480 MB per second. So one needs a PCI-X card or a PCI express x4 card. Secondly the card itself must be able to do 500 MB per second STR. Thirdly since the seek time of the HyperDrive is so small we need a very fast RAID card which does not degrade the seek time much. Fourthly the card must be able to process ideally 100,000 IOPS. The fastest of the lot in this regard is nvidia590 Mobo SATA RAID. This does not degrade the seek time of the HyperDrive4 at all since it uses the main CPU as a RAID processor. All other RAID solutions do degrade the seek time. But it will only stripe two disks effectively. The best PCI express solution we found was the Areca 1220 and its superior models. And the best PCI-X solution was the Silicon Image 3124 chip RAID solution. We used the Datoptic 4 external SATA card for these tests. It was the fastest card for 4 HyperDrives. Do not bother using the Silicon Image 3132 chip cards or any Marvell RAID solutions or any HighPoint RocketRaid cards. They are all too slow. We used an ASUS M2N32 WS Pro workstation board, a 2.61GHz dual core Athlon CPU, 1GB of 667MHz DDR2, a crummy GeForce 7100 Graphics card, Windows XP Service Pack2, the latest SATARAID5 driver for the 3124 chip from Silicon Image and the following Benchmarks... IOMeter from Intel HDTach3 from Simpli Software (for the STR) H2BenchW from Computer Technik Magazine The bottom line is as follows... This Benchmark represents a typical day in a high end users life with a mixture of file copying, software installation, virus scan. Word, Photoshop and swap file operation. The net result is that one HyperDrive4 unit is 524% faster than one Raptor. 2 RAID0 HyperDrive4s are 677% faster than 2 RAID0 Raptors and 4 RAID0 HyperDrive4s are 675% faster than any number of RAID6 Raptors aged with a 10% performance hit for defragmentation (42.5 score). The HyperDrive, being a Random Access Device, does not suffer noticeably from fragmentation. 4 RAID0 HyperDrive4 units have an STR of 394.7 MB per second (as measured by HDTach) and 386.6 MB/s (as measured by CT H2BenchW - which is the most realistic) and 445 MB/s (as measured by IOMeter, which must actually occur in some sense but is not so realistic). It has a max IOPS of 75,000 and a seek time of around 10 microseconds. The extra 9 microseconds over the 1 microsecond seek time of one unit are due to the processing time of the 3124 chip. RAID5 or RAID6 are unnecessary and counter-productive with HyperDrives, since the time needed for the RAID algorithm is too long for the seek time of the device. You would need a mainframe just for the

RAID algorithm! If you are concerned about memory stick failure which is a lot less likely than a mechanical failure in a hard disk, then use RAID10. Mirroring does not slow down the performance of the array. Let's state the conclusion one more time... 4 RAID0 HyperDrive4s are 675% faster than any number of WD740ADFD Raptors in RAID6 aged for defragmentation. If one made a 10 drive RAID6 array from 15,000 rpm SAS Seagate Savvios using the fastest Areca RAID card (ARC 1280) or the fastest Adaptec RAID card (AAR 4800SA) one could at best do 33% better that the 10 RAID6 Raptors on the Areca 1220 card i.e. get a score of 56.5. Then the 4 RAID0 HyperDrive array would be 508% better than any number of RAID Savvios. Here are the results of the tests we did in more detail for your perusal. These tests are really just stating the obvious in full technical detail...

Conclusion Instant computing is here. Any computer running on any number of hard disks is not a high performance machine. The bar has just been raised by 500% at a cost of 2000 ($4000). Copyright HyperOs Systems London March 2007