1. IBM FlashSystems Update and Marketplace Overview December 2013
John Sing
Executive IT Consultant
IBM Systems and Technology

2. Client Feedback: Data Economics
Business Critical…
“Keep me running!
Secure my data.”
Economics = Business Impact
Start Quickly… Add Value…
Data Intensive Business Solutions
“Can you get me up and running fast?”
“Can I add functions and scale?”
“How can I leverage all this data for my business and keep what I truly need?”
Economics =
Speed, Flexibility & Labor Costs
Economics =
LOB specific (Risk, Expertise, etc.) 2

3. Why Flash… In the last 10 years CPU Performance 8 - 10x increase
DRAM Speed 7- 9x
Network Speed 100x
Bus Speed 20x
Disk Speed 1.2x
…and everything waits
Source: IBM and Industry Estimates

4. The Economics of Flash 83% Overall lower software license costs
“I've taken away about four days of work time…down to about half an hour now. I'm doing it at the same cost I would have done four years ago, or probably lower.”
Reduced latency tenfold for Oracle Retail platform.
…5TB in 2U of rack space vs. 1,300 disk drives providing 400K IOPs, less than 1/10th the cost…
83% Overall lower software license costs
Due to fewer cores
Lower software maintenance
Much better storage utilization
As much as 50%
Lower maintenance
Ease management by 50%
38% Fewer Servers
Fewer cores
Fewer network connections
Environmentals 82% Lower Cost
Lower power / cooling
Less floor space

5. Why does Flash Latency make such a difference?
From an application’s perspective, all client requests take time…
I/O Time
Network Time
CPU Time
Newer CPUs help…
I/O Time
Network Time
CPU Time
But what if you could fix the I/O bottleneck?
I/O Time
Network Time
CPU Time
Time Recovered
Query Time
Hardware refreshes are done to lower the amount of time clients wait on data services. Ideally, the first refreshes come where the most of the client’s time is spent. The common assumption is CPUs are slow to process when most cases show that CPUs are waiting to get the data to process. Refreshing to the latest CPU technology helps, but addressing only CPUs only addresses the time CPUs process. This does not affect the majority of the client’s time waiting spent waiting on the disk.
By coupling servers with FlashSystem, customers are able to achieve a much greater $$/Gain as not only do clients achieve much higher performance, but we are able to consolidate and reduce expensive software licensing.
This approach benefits all applications and does not focus solely on the scalability of the applications. This addresses both scalability and the immediate issues of clients waiting. 5

6. CPU Utilization = ~4% FlashSystem CPU Utilization = ~50%
I/O Serviced by Disk
Issue/Process I/O request ~ μs
Wait for I/O to be serviced ~ 5,000 μs
CPU Utilization = ~4%
Less Work, More Wait
time
CPU Utilization
Work
Waiting for I/O
Work
Waiting for I/O
Work
Waiting for I/O
FlashSystem
CPU Utilization = ~50%
MORE Work, LESS Wait
I/O Serviced by FlashSystem
Issue/Process I/O request ~ μs
Wait for I/O to be serviced ~ 200 μs
time
CPU Utilization
Work
Wait
Work
Wait
Work
Wait
Work
Wait
Work
Wait
Work
Wait

7. Technology shift among storage tiers
Solid-State Drives (SSD)
Primary data
Cost-optimized
blended with
Nearline disks
Backup data
Disk and VTL
High capacitive Tapes
Long term archiving, “Cold data“
Cost-optimized
w/compression & deduplication
Linear Tape File System (LTFS)
Workflow
Data shipping

8. “All Flash” Systems from IBM
Redefine the Economics of Computing to Meet the Demands of a Smarter Planet
IBM announces new investments to help clients benefit from the value of flash
Investing ~$1B in flash system research & development
12 IBM Flash Centers of Competency around the world to help clients transform their systems
IBM integrates the IBM FlashSystemTM in the portfolio of IBM storage offerings

9. IBM Flash Storage Impact on Systems Economics
Better Economics Without Re-architecting Applications
85% Reduction
in batch processing times
50% Reduction
in Software Licenses
75% Reduction
in footprint; 1 Petabyte on 1 floor time
80% Reduction
in Energy Usage
100 µs Latency
No more bottlenecks
Enterprise Reliability
High Availability, 2D Flash RAID & IBM Variable Stripe RAIDTM
The data below are based on average operating conditions that may or may not be representative of a particular customer’s operating environment. The use case measurements are from TMS customers using the flash technology that has been integrated into IBM’s systems

10. Flash Enables a New, Optimized Infrastructure
for Operational, Transactional Workloads…
Core Transactional Systems
Social, Mobile, Analytics
Systems of Record
Systems of Engagement
System z
& Power
Workloads…
DB2
Secure Transactions
with Integrity
OpenStack, Hadoop…
2x Latency Improvement
5x Capacity Reduction
Rack Servers
Storwize with All Flash
IBM FlashSystem
Add Flash non-disruptively via DS8870 and Storwize SVC
Source: IBM Estimates

11. Journey to “Systems of Engagement”
Value: engage users in new, data-driven processes

12. Storage Spectrum with Flash
IBM FlashSystem flash storage system
100 microSeconds
Flash
Flash Drawers & Adapters High IOPs Performance
< 1 milliSecond
Solid State Disk
I/O performance in HDD form factor
Latency
IOPS
‘Hybrid’ HDDs with flash Improved I/O with spinning media capacities
Disk
5-15 milliSeconds
15k SAS Hard Disks
High performance spinning media
7.2k SAS Hard Disks
Capacity / Cost
Seconds
Tape
Backup, recovery, and archiving
Performance characteristics moving up from disk to flash
Cost / GB Flash $

13. Optimized IBM Flash Portfolio
Software
Servers
Storage
Flash in Server
EXP30 Ultra SSD Drawer
Flash Adapters, SSDs, EXP30 or EXP24S Direct Attached Drawers
Flash in Storage
IBM FlashSystem
Flash Optimized Systems, All-Flash, Hybrid Flash,
Flash Caching
IBM High IOPS Adapters + FlashCache
Flash Server Caching w/ Flash Adapters and SSDs
Flash Cooperative Caching
EXP30 Ultra SSD Drawer + IBM DS8870 Storage System + Easy Tier Server
Advanced Tiering Software w/ Flash Drawers, Adapters, SSDs
Bottom graphic…benefits equal cooperative and shared storage.
Flash

14. IBM Flash Software IBM FlashCache Storage Accelerator (IBM System x)
5th Generation Easy Tier
(IBM DS8870 with IBM Power 7)
Disks
Server Cache
Close storage to CPU performance gap
Tier data and boost storage performance with Easy Tier Server
IBM FlashCache Storage Accelerator -
Increase storage performance to close storage to CPU gap
Use best mix of Flash and hard drives for price / performance
Add Flash to match the desired performance level and size of the hot spot
Most beneficial for read I/O
Write I/O performance improves as a side effect of eliminating read I/O
Best cost per I/O to achieve performance levels
Offers out of the box application acceleration (no tuning or data placement required)
Pre-sales profiler tool will help clients purchase only the Flash they need
IBM DS8870 with 5th Generation Easy Tier –
Tiering data to the application server will boost performance
Easy Tier moves the hottest data closer to server and overcomes latency of the SAN
Both server and storage resources remain optimized for performance and cost objectives
Advanced self-tuning reduces administration costs further and allows IT staff to focus on strategic initiatives
Preserves advanced data protection / HA solutions

15. EasyTier + SSDs in Servers
More Tiers APIs to Software
Cluster with
Direct Attach
SSDs
Combined DAS and External Storage
With Data Coherency
....
SSD
SSD
Strong Performance
Weak Storage Function
....
SSD
SSD
....
SSD
SSD in
SAN/NAS Storage
SAN/NAS
Shared Storage
SSD
Strong Storage Function
Not Optimal Performance
Strong Storage Function
And Performance
Today 2 ways of solving the memory speed storage workload problem. Create scalability, integrity, reliability and availability problem.
EasyTier extended to Server SSD - a unified scale out and scale up storage platform solving mission critical memory speed storage workload problem.

16. IBM Enterprise-Class Flash in DS8870
Enterprise Storage for Critical Applications
Superior Reliability
Unique Server Integration
Exceptional Performance
Efficiency & Optimization

17. Optimized Local Data Placement – Easy Tier
System learns workload, does automated relocation, removes user complexity
Easy Tier for DS8000, SVC, V7000
V1, V Tier
Now Tier
New Server attached Flash Tier
Future LTFS Tape as a Tier
SSD RAID Array(s)
FC/SAS RAID Array(s)
SATA RAID Array(s)
SSD
NL SAS
Logical volume
VTS
LTFS
SAS HDD
Easy Tier can provide better performance in 1/2 the floor space
40% less energy consumed
96 NLSAS HDDs + 16 SSDs (5.7 KW) vs. 192 FC HDDs (9.5 KW)
This is an SPC-1 Benchmark we submitted in April for the DS8700 and Easy Tier, and it is the first (and only) SPC-1 benchmark to include SATA drives. This demonstrates that smart data placement can enable tier-one performance with a combination of SATA and SSDs.
You can see that by moving just a small portion of the data from the HDDs to the SSDs can address performance requirements without the excessive costs of putting ALL the data on expensive SDDs. The graph shows how after Easy Tier automates the movement of hot data to the SSDs, performance has improved by over 5X … with no user intervention!
3-4x DB2 throughput improvement on brokerage application. Configuration:
POWER7 (Model 9117-MMB), DB2 9.7 FP1, DS8700
4.6 TB SSD TB 15K FC HDD)
Ds8700 with Easy Tier achieves better performance in half the floor space and with 40% less energy consumption. Single frame Ds8700 with 96 SATAs + 16 SSDs (5.7KW) vs. Dual frame Ds8700 with 192 FC HDD (9.5KW)
Futures tier architecture support, single tier support, application assisted support, power mgmt support

18. Easy Tier V4 Application Performance
Equivalent performance with or without encryption drives in the DS8800
As encryption is done at the drive level, there is no performance impact meaning the performance of an encryption drive or a non-encryption drive is the same. In this example, a DB2 Brokerage transactional workload is utilized and clearly demonstrates this important point and differentiates the DS8000’s implementation of encryption from our competitors.
Configuration
96 146GB 15K drives (FDE or non-FDE) – RAID-5
16 400GB SSDs (FDE or non-FDE) – RAID-5
2 DS8800, 1 without encryption with non-FDE drives, 1 with encryption with FDE drives
DB2 9.7 FP1, 2 Instances, 2 x 2 TB DBs, 2 Buffer Pools at 54 GB each
2 x 2 TB volumes were allocated for database, temp files and data generation
2 x 100 GB volumes were allocated for log files

19. Proven enterprise flash for DS8870
All-flash DS8870 delivers the ultimate in performance, resiliency and economics
Same usable capacity
Much better performance
Much smaller footprint
Much less energy
And can reduce overall IT infrastructure costs
1,056 15k hard disk drives
224 flash drives
146GB 15k
(RAID 10)
All-Flash (RAID 5)
All-Flash Benefit
Raw Capacity
152TB
90TB
41% Less
Usable Capacity
72TB
Same
Response Time
~ 1 ms
~ 0.3 ms
70% Less
Drive Count
1,056
224
80% Less
Frames
3 frames
2 frames
33% Less
Energy Usage
13.9kw
5.3kw
62% Less
$ / GB / IOP
Equivalent NA
Source: Wikibon
All-HDD
All-flash or hybrid-flash
Performance based on measurements and projections using IBM benchmarks in a controlled environment.

20. Start Small and Grow – Storwize Platform
Start Small & Scale-up
Rapid
Deployment
Buy Less
Storage
Use What
You Have
IBM & non-IBM
Built-in
Expertise
Simple, Rapid Development
IBM Storwize Platform
Converting SVC from a Product to a Platform
Real-time Storage Usage Analytics
Over 36,000 systems shipped in 2 years since inception…
Over $1B revenue in 24 months…
Award winning storage platform: features, quality, reliability, support…
Packages
IBM Storwize SVC (Software Appliance; No Storage)
IBM Storwize 3K (Low-end)
IBM Storwize 7K (Larger)
Priced Options
IBM Storwize Real-time Compression
IBM Storwize Unified (File)
IBM Storwize Metro / Global Mirror
IBM Storwize External Virtualization
Under Consideration
IBM Storwize Flash
IBM Storwize ObjectStor
IBM Storwize Cloud
Embedded Options
IBM PureSystems
Business Partner Solutions
…powered by Storwize embedded…

21. IBM FlashSystem Solution
Combining IBM SVC & IBM FlashSystem™
In a 24/7 market, enterprise clients often strive to achieve maximum performance coupled with deep features and functionality. And naturally, it must be delivered cost effectively. With the introduction of the IBM SAN Volume Controller and IBM FlashSystem bundle, a solution now exists
IBM clients gain…
The extreme performance of IBM FlashSystem with IBM MicroLatency™
Advanced storage functionality of IBM SVC
Thin Provisioning – allocate storage “just in time”
Easy Tier – storage efficiency
FlashCopy – point in time copies
Mirroring/Copy Services – data replication and protection
Real-Time Compression – up to 5X more data in the same physical space
An ability to cost effectively deploy quickly and realize immediate results

22. IBM Storwize V7000 FlashSystem
Extreme Performance
Feature Rich SDS Systems
with
IBM Storwize V7000 FlashSystem Edition
Competitive Economics
Over 250K IOPS
Flash Copy for Backup & Optimal Workload Availability
$/TB Value with Thin Provisioning & Real Time Compression
Storage Efficiency with Easy Tier
IBM FlashSystem 820
Extreme Performance with IBM MicroLatency™
All Flash 20TB RAIDed data capacity
Macro Efficiency 1U form factor
Variable Stripe RAID™ + 2D Flash RAID for Enterprise Reliability
V7K + SSDs: use for up to a couple of TBs of "fast" capacity and less than 100K IOPS. Great entry point into flash capabilities.
V7K + FlashSystem: use for 1-2 FlashSystems worth of "fast" capacity and over 250K IOPS (need some actual testing to determine the ceiling). Great expansion play once customer has tasted flash.
FlashSystem + SVC: use for everything above that. Deployment at scale. 22

23. XIV SSD Solution High capacity SSDs used as secondary cache
400GB SSD device per module
6TB of total cache per rack
Housed in PCI caddy in rear of module
Added to all installed modules
No tuning necessary
Immediate gain for all applications
Optional selective volume activation
Available to all Gen3 systems
Order with SSD from factory
Field non-disruptive upgrade
In this page we we describe a bit more detail of the SSD physical implementation:
Here you can see a simple diagram of the back of an XIV Gen3 Module.
It is basically 1 PCI slot. One per Module and up to 15 in a fully loaded XIV Grid.
The SSD disks are 400GB MLC, which translates to a total of 6TB per RACK.
Performance is gained as soon as the disk is phased-in. No tuning, Zero Configuration required.
Customers have two ways to acquire this solution:
A: From Factory. Already inserted.
B: As a non-disruptive field upgrade. Which applies to all existing XIV Gen3 Customers.

24. Outstanding Applications Performance
With SSD Caching
DB2 Brokerage (IOPS)
Heavy Random Brokerage
90/10, Mixed block IO
84% Random Read Miss
WebSphere Datastore (IOPS)
Web 2.0 OLTP Workload
80/20/4k
Core ERP (IOPS)
CRM and Financial DB Workload
70/30/8k
Medical Record App Server (RT)
Healthcare EMR Workload
100% random IO
Performance is based on measurements and projections using standard IBM benchmarks in a controlled environment. The actual throughput that any user will experience will vary depending upon considerations such as the amount of multiprogramming in the user's job stream, the I/O configuration, the storage configuration, and the workload processed. Therefore, no assurance can be given that an individual user will achieve throughput improvements equivalent to the performance ratios stated here.

25. Completing IBM storage portfolio with All-Flash Arrays
Accelerate critical applications and unleash the power of performance
Direct Attach
All Flash Arrays
Hybrid Storage Systems
IBM FlashSystem All-Flash Arrays offer extreme performance, low latency and enterprise attributes to handle fast moving operational data in real time
Direct Attach Flash (PCIe adapters, server SSDs) provides low latency, high performance storage for in-server applications
Hybrid Storage Systems (Storage with SSDs & HDDs) provide a higher level of enterprise capabilities with superior capacity

26. Hybrid arrays vs. dedicated Flash arrays?
What you get:
A consolidation play
Non-deterministic performance
Ease of integration
Licensing (still common)
Built-in Tiering
Hot-spots
Another tier to manage
Relative ease of use
Marginal application efficiencies
Proven resiliency
SSDs
What you don’t get:
Best Latency / Best Performance
Deterministic/Consistent Performance
Architecture design for Flash
Reduced footprint
Rack space reductions
Power/Cooling savings
Simplicity
Lesser Tiers
Zero Tuning
Maximum application efficiency

27. IBM FlashSystem Portfolio
Extreme Performance
MicroLatency™
Macro Efficiency
Enterprise Reliability
Extreme Performance
Boost enterprise productivity
Improve application performance
Ideal for OLTP, data warehousing and OLAP environments
High availability system
IBM FlashSystem 820
IBM FlashSystem 720
IBM FlashSystem 810
IBM FlashSystem 710
IBM FlashSystem Solution
IBM FlashSystem 820
San Volume Controller (SVC)
Enterprise Features
Acceleration of applications
Enterprise feature sets to ensure data protection
There is a need for extreme raw performance of low latency, IOPS and bandwidth

28. IBM FlashSystem 720 / FlashSystem 820 Architecture
Redundant Management Control Processors
Redundant Power Supplies
Two Dual-Ported 8 Gb/s FC or 40 Gb/s QDR IB Interfaces
N+1 batteries
Redundant
RAID controllers
12 Flash Modules
(10+1+1)
1U Chassis
Redundant Fans

29. 1 Petabyte, 1 Rack, 22 Million IOPS
What it means… Or, solving for 22 Million IOPs…
Either 315 Racks
Performance Optimized Disk…
IBM FlashSystem
1 Rack
1 Petabyte: 1 Floor Tile
100 microsecond latency
22 Million IOPS
210 GB/s
12.6 KW power
Or 630 Racks
High Capacity Disk…
Less power than the average 200TB array
Source: IBM Actual Results and Estimates I

30. Tipping Point Demonstration
IBM FlashSystem, IBM Power Systems and DB2
10 GbE
Networking
Highly Scalable & I/O Intensive
OLTP Database Workload
Compelling Economics Significantly improved workload efficiency
Extreme Capacity Buy only what you need; add capacity as needed
Application Transparency Avoid risk and cost of change as you grow or subside
Continuous Availability Uninterrupted access to data with consistent performance
Fibre Channel
Networking
IBM FlashSystem 820
(4-1U units, 20TB Each)
IBM Power 780
(4 nodes, 128 Cores,
2TB Memory)
IBM DB2 v10.5
(10-8 core cluster members)
Improve Performance
Reduce Costs
Enable New Opportunities
Leverage the “Economies of Scale” of Flash
Accelerate Application Performance
Gain Greater System Utilization
Lower Software & Hardware Cost
Save Power / Cooling / Floor Space
Drive Value Out of Operational Data
ibm.com\storage\flash 30

31. Tipping Point Demonstration Results
Normalized $ / IOPS
1.3 Million IOPS
43K+ Transactions p/sec
13K Updates per second
11x Less
IBM
FlashSystem
2,500Spindles+ 128 SSDs
5,000 Spindles
Energy
Space
26x Less
80x Less
IBM
FlashSystem
2,500Spindles+ 128 SSDs
5,000 Spindles
IBM
FlashSystem
2,500Spindles+ 128 SSDs
5,000 Spindles
IBM FlashSystem
Cost: $xxx
Option #1
2, ” Disks SSD’s
8 Racks, 36.79kVa; 114,700 Btu/hr
Annual Power Cost: $88,198
Storage Cost: 10.6M (List Price)
Compare #2
5,000 Disks
18 Racks, 75.64kVa; 229,400 Btu/hr
Annual Power Cost: $177,232
Storage Cost: 13.6M (List Price)
ibm.com\storage\flash 31

32. IBM Differentiators – Enterprise Reliability
Superior Durability: Using the Best Flash
10X
30X
P/E* Cycles
*Program/Erase (Write)
Flash Type
MLC
eMLC
SLC 0K
100K
50K
Superior Protection: Beyond Disk RAID
Variable Stripe Sizes
Read Disturb Mitigation
Automatic Read Sweeper
High-Speed Clock Recovery
Advanced Engineering = Less Maintenance
Protection Within And Across Flash Modules
Self-Recovering Flash Modules
Avoid system rebuilds

33. IBM FlashSystem Roadmap
2013
2014
2015+
IBM FlashSystem 710, 810, 720, 820
1U, 24TB,
32nm SLC, eMLC
450K IOPS,
8Gb FC, QDR IB  
Next Generation FlashSystem
Next Generation F/O FlashSystem
FlashSystem Solution
5U 24TB
Scalable to 8x SVC FlashSystem 820
SVC Integration
Enhancements
SVC Integrated
Management
Next Gen SVC Integrated Mgmt / Adv. Function
PureSystem Integration
Next Gen PureSystem
FlashSystem F/O Flex Node
Legend
Integration into Big Data/ Analytics solutions
Next Gen. Big Data/Analytics solutions DS8K and Storwize integration
Available
Future
Dashed plan means finalized project plans are not in place: hardware, people, deliverables defined, scoped, and milestones defined. 33 33

34. IBM FlashSystem: A Marketplace Overview
So many vendors, so little time……

35. A World Full of “Flash” Options
FlashSystem

36. Everyone uses Flash The difference is how it is implemented
HW-based Vs SW-based architecture
Purposed-built for flash or not?
Flash Chip Choices
Data Protection Schemes
Scalability, Reliability, Availability
Method of Deployment
Functionality and Features
Hybrid or All Flash
And yes… Latency!
IBM FlashSystem
HW-only data path
FPGA

37. SSD is not Flash Storage
Flash inside a Disk Enclosure
Technology designed for Disk, not Flash
Data protection outside SSD
Data Path handled by OS/Controllers outside SSD
Performance is reduced and marginal
Added Latency by:
Controllers
SW Layers
SAS Controllers HDDs
Tiering
Shared data path
SSD
Common in hybrid arrays
Focus is to maintain functionality and consolidation
Focus on Tiering and Data Movement
SSD competes with all other disks’ for resources
When performance matters (which is ALWAYS), SSD is not the best

38. SSD is not Flash Storage
Rule of Thumb: consider 1U IBM FlashSystem when > 5 SSD
SSD
525K IOPS 300 Watt 25/100 µs write/read
FlashSystem 820 FlashSystem 720
EasyTier
(SSD and disk mix)
Why? IBM FlashSystem provides better lifetime, higher performance and smaller energy foot-print when needing more Flash than 5 SSDs
SSD when:
< 6 SSDs for EasyTier
SVC (external EasyTier) not indicated solution

39. IBM FlashSystem Focus areas
Performance (Latency):
HW-based design/FPGA
Efficiency:
Application/Server, Data Center
Reliability:
VSR, 2D Flash RAID

40. Flash HW and SW-based designs
SW-Based Designs: SW Controlled Data Path
Front End Directors
Back End Directors OS
CPU
Outside World
SSD/SAS/SATA HDD’s
Cache
Added latency by:
Array Controllers and SW Layers
SAS Controllers and Shared bus
Tiering and Shared data path
Form Factor Enclosure
SSD

41. Flash HW and SW-based designs (cont)
HW-Based Designs: HW controlled data path
Maximum performance and lowest latency by:
Optimized FPGA HW data path
Custom HW design
Massive parallelism handling data
25us write latency into flash media
IO Interface FPGA
FPGA
CPU
Outside World
Flash Media
Autobahn

42. IBM FlashSystem Core Performance Concepts
IBM FlashSystem is a hardware-only data path
Custom FPGA-based data movement decreases latency vs. software
Focus is on MicroLatencyTM on standard SAN interfaces vs. competitors
Comparable latency on SAN compared to DAS (PCIe)
All the benefits of shared storage with no latency impact
Distributed out-of-data-path CPU processing
16 PPC CPUs
IO Modules
RAID Controllers
12x
Flash Modules

43. Positioning IBM FlashSystem in a broad marketplace
<100us
1M+ IOPS
<300us
500K IOPS
FlashSystem
<1ms r-latency
100K IOPS
<5ms r-latency
100K IOPS

44. Change the Economics of Storage
Focus on those that deliver all of these, and more…
Reduced TCO
Business Value
Reduce Latency
Increase Manageability
Cut Environmental Costs
Improve Compute Efficiency
Guarantee Reliability
Provide Scalability

45. IBM Flash Storage Portfolio Meeting the Demands of a Smarter Planet
Flash in
IBM XIV
and DS8870
All-Flash Storwize V7000
IBM Flash Drawers
Adapters & SSDs
IBM FlashSystem
IBM Flash Software for Management, Caching & Tiering
Our acquisition of TMS is not where we begin this journey…TMS is how we continue to round out a portfolio that has shipped over 72PB of flash to date. We are firmly focused on taking the entire IBM technology offering and accelerating to unprecedented levels of performance.
All of the above supported with Flash Storage TODAY! 45

46. Thank You