SF 2009
Future Solid-State Drive (SSD) Innovations Knut Grimsrud Intel Fellow, Director of Storage Architecture Chris Saleski Intel Business Development Manager
MEMS004
Agenda Infrastructure Improvements SSD Improvements New Storage Interfaces
2
Driver & Stack Efficiencies Read Latency & Transfer Rate
• Driver & stack optimization targets change with SSDs
• Time spent • Time saved
= 100us = 2.5ms*5% = 125us
0.300
Service Time (ms)
– For HDD, extra 100us overhead is good tradeoff if it saves 50% on 5% of IOPS that take 5ms each
Total overhead at intercept (~60us)
0.350
0.250 0.200
~2.5X IOPS impact
0.150
Transfer rate is inverse of slope (~285MB/s)
0.100 0.050
– For SSD, this would be a bad tradeoff • Time saved negligible and does not offset time spent
0.000 0
4
8
12
16
20
24
28
Transfer Size (KB) OS A, Driver X
OS B, Driver X
OS B, Driver Y
Driver and stack efficiency optimizations need to focus on latency to best realize SSD performance potential
3
Performance measurements are made using specific computer systems and/or components and reflect the approximate performance of the technology as measured by those tests. Any difference in system hardware or software design or configuration may affect actual results.
32
What’s 100us Among Friends? PCMark* Vantage HDD Relative Performance Intel® X25-M SSD
HDD
0
5000
10000
15000
20000
25000
Benchmark Score (higher is better) Driver Y
Driver X
Driver efficiencies might have modest effect on HDD performance, but can have substantial impact on SSDs • Other names and brands may be claimed as the property of others For System configuration A – see backup
4
Performance measurements are made using specific computer systems and/or components and reflect the approximate performance of the technology as measured by those tests. Any difference in system hardware or software design or configuration may affect actual results.
Controller Efficiencies 4-Drive Intel® X25-E SSD Config Random Read Performance 180,000
A Ports
160,000 140,000
PCIe
ESB
STP
SAS
Total IOPS
IOH
SATA
B Ports
120,000 100,000 80,000
~3.5X IOPS impact
60,000 40,000 20,000 0 512
1K
2K
4K
8K
16K
32K
Transfer Size
Controller designs need to be re-examined to get the most out of the concentrated performance SSDs provide System configuration B – see backup
5
Performance measurements are made using specific computer systems and/or components and reflect the approximate performance of the technology as measured by those tests. Any difference in system hardware or software design or configuration may affect actual results.
64K
6
7
OS Tailoring for SSD – TRIM Relative PCMark* Vantage HDD Scores w/ Intel® X25-M SSD
30000 25000 20000 15000
2% )
y ed
(8
pt TR IM m
Em 1%
pt y
20 %
Em
pt y
40 %
Em
pt y
60 %
Em
pt y Em
80 %
(8 ti n e Pr is
Note expanded scale
2%
)
10000
Infrastructure accommodations for SSD-specific behaviors provide visible end-user benefits Filled region of drive filled with random 4KB writes over fill space System configuration A – see backup
8
Performance measurements are made using specific computer systems and/or components and reflect the approximate performance of the technology as measured by those tests. Any difference in system hardware or software design or configuration may affect actual results.
Intel® SSD Toolbox Supports Range of OS’s
Although some OS’s will provide native TRIM support, capability for TRIM support with other OS’s also feasible 9
Prefetching Efficiencies
Value normalized to w/ prefetching
• Since some prefetching approaches may be based on HDD behavior assumptions, prefetching can be unhelpful for SSDs 100% 90% 80% 70% 60% 50%
3.0X
40%
2.7X
30% 20% 10% 0% w/ Prefetching
w/o Prefetching
Benchmark* Score
w/ Prefetching
w/o Prefetching
I/O Operations
w/ Prefetching
w/o Prefetching
MB Transferred
Disabling prefetching with SSDs can have platform benefits in terms of both performance and power *Data captured from PCMark Vantage system benchark suite System configuration A – see backup
10
Performance measurements are made using specific computer systems and/or components and reflect the approximate performance of the technology as measured by those tests. Any difference in system hardware or software design or configuration may affect actual results.
Score Read Write
Agenda Infrastructure Improvements SSD Improvements New Storage Interfaces
11
12
13
14
15
16
Short Term Performance Variability • Average performance (even over time) does not paint complete picture – Short term performance variability also highly impactful Performance Variability 160
Throughput (MB/s)
140 120 100 Read MB/s Write MB/s
80 60 40 20 0 0
20
40
60
80
100
120
Time (minutes)
System configuration D – see backup
17
Performance measurements are made using specific computer systems and/or components and reflect the approximate performance of the technology as measured by those tests. Any difference in system hardware or software design or configuration may affect actual results.
Improving Short Term Variability 4K Random Write IOPS Variability
IOPS Histogram
40000
40000
35000
35000
30000
30000
25000
25000
20000
20000
15000
15000
10000
10000
5000
5000 0
0
~20X variability 0
5
10
15
20
25
30
100%
80%
60%
40%
20%
0%
Time (Minutes)
Performance variability will substantially improve and increased focus on the minimum performance rather than averages/peaks System configuration C – see backup
18
Performance measurements are made using specific computer systems and/or components and reflect the approximate performance of the technology as measured by those tests. Any difference in system hardware or software design or configuration may affect actual results.
IOPS
IOPS
~25% variability
Traditional Caching Protection • Protecting the volatile write buffer required for high-integrity environments • Some approaches can be cumbersome, expensive, and prone to reliability issues *
19
* Other names and brands may be claimed as the property of others
2X 4.7uF PowerStor* 2.5V caps
More Efficient Solution & Approach 1.3 V
• Intel® SSD technology has very modest amount of data buffered
3.3 V
3.3 V
SDRAM
• Because of very modest amount of buffering, the energy required to reliably save buffered data is quite small
SoC 19x19
– Host data does not go via the DRAM
– Only a few modest-sized caps required to make buffer nonvolatile
Solution so efficient that drives with protected write buffers may not be readily discernible
20
21
Buffer Protection Behavior & Results Powerloss detect signal triggers 5V power lost at this point
Slope at this point determines expected hold-up time while active
Ch Ch Ch Ch
1: 2: 3: 4:
Buffer flush handler entered Power Fail Int# 3.3V to the NAND Handler busy 5V
Slope at this point not representative since flush finished and everything shut off
5V voltage drops low enough for 3.3V regulator to droop
This is NOT the keep-up time
Full worst-case characterization required to ensure reliability for all cases. Keep-alive time not as long as implied. 22
Summary • Infrastructure improvements – Driver and stack efficiencies have substantial impact on effective SSD performance – Controller efficiencies also have major impact – Platform IOPS scaling improvements on the horizon to better utilize concentrated SSD performance – TRIM provides user-visible benefits – Prefetching can get in the way for SSDs
• SSD improvements – Sustained performance stability will see major improvements – Short-term performance variability will be addressed – Efficient solutions for protecting volatile buffers will be available to better address needs of high-integrity environments
23
Agenda Infrastructure Improvements SSD Improvements New Storage Interfaces
24
Motivation for New Storage Interface in Servers 120000 100000 IOH
1 SATA
PCIe STP
IOPs
2
80000 60000 40000
ESB
SAS
20000
1
2
0 SAS RAID controller
Sata Host Raid
Hardware Attach Point Affects Performance 25
Motivation for New Storage Interfaces (Cont’d) • High performance platform I/O connection point
•
Bandwidth:
• SATA2 - 0.3GB/s (SATA3 – 0.6GB/s) • x4PCIe2 – 2.0GB/s • Bandwidth scalability • 0.5GB/s per lane up to 8GB/s
•
Latency
• 4kbyte SATA2 bus transit adds ~20usecs
•
Host management possible
─ Faster RAID controllers could bridge the SATA/SAS ─ Storage infrastructure un-established ─ Some servers have few PCIe slots
bottleneck
PCIe – an Interesting Attach Point for Some I/O Intensive Applications 26
PCIe SSD Technology Demo
Server Platform CPU
Intel S5520HC Two Intel® Xeon® Processor 5500 Series
Memory
12GB DDR3 1333Mhz
OS
Microsoft Windows* Server 2008 Standard
Storage
(7) 300GB PCIe x8 Intel Prototype SSDs
Break-through Sustained IOPs Running OLTP 27
*Other names and brands may be claimed as the property of others
Motivation for New Storage Interface in Clients • Low local capacity requirements in IT “cloud model” – Can be satisfied with SFF SSD or PCIe
• Lower TCO with future SSDs + + + + +
28
Higher performance Lower than HDD floor cost Small form factor Light weight Energy efficient
A new Logical SSD Interface: Enterprise NVMHCI •
PCIe SSDs lack standard OS support & driver infrastructure, since there is no standard host controller interface
•
Impact of no standard infrastructure:
•
Enable broad adoption by extending NVMHCI (Non-Volatile Memory Host Controller Interface) to Enterprise
– Requires SSDs vendor to provide drivers for every OS – OEM features, like error logs, are implemented in an inconsistent fashion – OEMs have to validate multiple drivers
– Leverage the client NVMHCI i/f, software infrastructure, and Workgroup to fill gap quickly with streamlined solution
Enterprise NVMHCI brings standard infrastructure to PCIe SSDs 29
Summary • Hardware Attach Point Makes a Difference • Break-through Sustained IOPs Running OLTP • Future (IT) Client SSDs: faster boot, Lower Power, Small Form Factor, Lighter Weight • Enterprise NVMHCI brings standard infrastructure to PCIe SSDs
30
Want More Info on SSDs? • Attend or download these SSD-related sessions
Tuesday, Sept 22nd – EBLS001 - Extending Battery Life of Mobile PCs: An Overview Wednesday, Sept 23rd – MEMS001 - Designing Solid-State Drives into Data Center Solutions – MEMS002 - Understanding the Performance of Solid-State Drives in the Enterprise – MEMS003 - Enterprise Data Integrity and Increasing the Endurance of Your SolidState Drive – MEMS004 - Future Solid-State Drive Innovations – MEMQ002 - Open Q&A for SSD sessions Thursday, Sept 24th – MPTS006 - Extreme Notebook Design: Architecting the Most Powerful Mobile Platforms for Gaming & Workstation Applications – RESS006 - Differentiated Storage Services: Making the Most of Solid-State Drives – STOS004 - Intel® Modular Server with Intel® Solid-State Drives
• Visit our Booth #532 on Level 1 of the Tech Showcase – SSD vs HDD comparisons, gaming demo and more!
• Visit us online at www.intel.com/go/ssd
– Product briefs, datasheets, whitepapers, videos, technical support
31
Legal Disclaimer • INFORMATION IN THIS DOCUMENT IS PROVIDED IN CONNECTION WITH INTEL® PRODUCTS. NO LICENSE, EXPRESS OR IMPLIED, BY ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS DOCUMENT. EXCEPT AS PROVIDED IN INTEL’S TERMS AND CONDITIONS OF SALE FOR SUCH PRODUCTS, INTEL ASSUMES NO LIABILITY WHATSOEVER, AND INTEL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY, RELATING TO SALE AND/OR USE OF INTEL® PRODUCTS INCLUDING LIABILITY OR WARRANTIES RELATING TO FITNESS FOR A PARTICULAR PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT. INTEL PRODUCTS ARE NOT INTENDED FOR USE IN MEDICAL, LIFE SAVING, OR LIFE SUSTAINING APPLICATIONS. • Intel may make changes to specifications and product descriptions at any time, without notice. • All products, dates, and figures specified are preliminary based on current expectations, and are subject to change without notice. • Intel, processors, chipsets, and desktop boards may contain design defects or errors known as errata, which may cause the product to deviate from published specifications. Current characterized errata are available on request. • Nehalem and other code names featured are used internally within Intel to identify products that are in development and not yet publicly announced for release. Customers, licensees and other third parties are not authorized by Intel to use code names in advertising, promotion or marketing of any product or services and any such use of Intel's internal code names is at the sole risk of the user • Performance tests and ratings are measured using specific computer systems and/or components and reflect the approximate performance of Intel products as measured by those tests. Any difference in system hardware or software design or configuration may affect actual performance. • Intel, Core and the Intel logo are trademarks of Intel Corporation in the United States and other countries. • *Other names and brands may be claimed as the property of others. • Copyright © 2009 Intel Corporation.
32
Risk Factors The above statements and any others in this document that refer to plans and expectations for the third quarter, the year and the future are forward-looking statements that involve a number of risks and uncertainties. Many factors could affect Intel’s actual results, and variances from Intel’s current expectations regarding such factors could cause actual results to differ materially from those expressed in these forward-looking statements. Intel presently considers the following to be the important factors that could cause actual results to differ materially from the corporation’s expectations. Ongoing uncertainty in global economic conditions pose a risk to the overall economy as consumers and businesses may defer purchases in response to tighter credit and negative financial news, which could negatively affect product demand and other related matters. Consequently, demand could be different from Intel's expectations due to factors including changes in business and economic conditions, including conditions in the credit market that could affect consumer confidence; customer acceptance of Intel’s and competitors’ products; changes in customer order patterns including order cancellations; and changes in the level of inventory at customers. Intel operates in intensely competitive industries that are characterized by a high percentage of costs that are fixed or difficult to reduce in the short term and product demand that is highly variable and difficult to forecast. Additionally, Intel is in the process of transitioning to its next generation of products on 32nm process technology, and there could be execution issues associated with these changes, including product defects and errata along with lower than anticipated manufacturing yields. Revenue and the gross margin percentage are affected by the timing of new Intel product introductions and the demand for and market acceptance of Intel's products; actions taken by Intel's competitors, including product offerings and introductions, marketing programs and pricing pressures and Intel’s response to such actions; and Intel’s ability to respond quickly to technological developments and to incorporate new features into its products. The gross margin percentage could vary significantly from expectations based on changes in revenue levels; capacity utilization; start-up costs, including costs associated with the new 32nm process technology; variations in inventory valuation, including variations related to the timing of qualifying products for sale; excess or obsolete inventory; product mix and pricing; manufacturing yields; changes in unit costs; impairments of long-lived assets, including manufacturing, assembly/test and intangible assets; and the timing and execution of the manufacturing ramp and associated costs. Expenses, particularly certain marketing and compensation expenses, as well as restructuring and asset impairment charges, vary depending on the level of demand for Intel's products and the level of revenue and profits. The current financial stress affecting the banking system and financial markets and the going concern threats to investment banks and other financial institutions have resulted in a tightening in the credit markets, a reduced level of liquidity in many financial markets, and heightened volatility in fixed income, credit and equity markets. There could be a number of follow-on effects from the credit crisis on Intel’s business, including insolvency of key suppliers resulting in product delays; inability of customers to obtain credit to finance purchases of our products and/or customer insolvencies; counterparty failures negatively impacting our treasury operations; increased expense or inability to obtain short-term financing of Intel’s operations from the issuance of commercial paper; and increased impairments from the inability of investee companies to obtain financing. The majority of our non-marketable equity investment portfolio balance is concentrated in companies in the flash memory market segment, and declines in this market segment or changes in management’s plans with respect to our investments in this market segment could result in significant impairment charges, impacting restructuring charges as well as gains/losses on equity investments and interest and other. Intel's results could be impacted by adverse economic, social, political and physical/infrastructure conditions in countries where Intel, its customers or its suppliers operate, including military conflict and other security risks, natural disasters, infrastructure disruptions, health concerns and fluctuations in currency exchange rates. Intel's results could be affected by adverse effects associated with product defects and errata (deviations from published specifications), and by litigation or regulatory matters involving intellectual property, stockholder, consumer, antitrust and other issues, such as the litigation and regulatory matters described in Intel's SEC reports. A detailed discussion of these and other risk factors that could affect Intel’s results is included in Intel’s SEC filings, including the report on Form 10-Q for the quarter ended June 27, 2009. Rev. 8/3/09