VM24

2010 IBM Corporation
IBM Power Systems

October 2010
PowerVM vs VMware
Technical Comparison
IBM Power Systems
2
Virtualization Basics
Hypervisor
Server Hardware
Workload
?
Virtual Machines
Workload
D
Workload
C
Workload
B
Workload
A
IBM Power Systems
3
Agenda
Hardware
Hypervisor
Workloads
Customer References
TCA/TCO Examples
Summary
IBM Power Systems
44
Hardware
IBM Power Systems
5
Power Systems
TM
Hardware
Complete Line of Systems
POWER7 Processor Technology
Balanced System Design
Performance Leadership
Built-in Virtualization
Reliable Hardware
Energy Efficient
IBM Power Systems
6
POWER Processor Technology
2001
Dual Core
Chip Multi Processing
Distributed Switch
Shared L2
Dynamic LPARs (32)
2004
Dual Core
Enhanced Scaling
SMT
Distributed Switch +
Core Parallelism +
FP Performance +
Memory bandwidth +
Virtualization
2007
Dual Core
High Frequencies
Virtualization +
Memory Subsystem +
Altivec
Instruction Retry
Dyn Energy Mgmt
SMT +
Protection Keys
2010
Multi Core
On-Chip eDRAM
Power Optimized Cores
Mem Subsystem ++
SMT++
Reliability +
VSM & VSX (AltiVec)
Protection Keys+
POWER8
Concept Phase
POWER4
180 nm
POWER5
130 nm
POWER6
65 nm
POWER7
45 nm
IBM Power Systems
7
8 Sockets
100 GBps
50 GBps
24 MB SRAM
256 KB / Core
2.26 GHz
1 / 2
4 / 6 / 8
2.3 B
596 mm
2
45 nm Hi-k
(Nehalem-EX)
Intel Xeon 7560
4x
3.6x
2x
1.33x
~
1.7x
2x
~
.5x
~
~
32 Sockets Scalability
360 GBps SMP Fabric
1 / 2 / 4 Threads per Core
45 nm Cu, SOI, eDRAM Technology
100 GBps
32MB eDRAM
256 KB / Core
4 GHz
4 / 6 / 8
1.2 B
567 mm
2
POWER7
Maximum Frequency
L3 Cache
Memory Bandwidth
Cores
L2 Cache
Transistors
Size
Processor Comparison
IBM Power Systems
8
Balanced System Design
System
Advanced SMP links provide near
linear scaling for POWER7 systems
Increased Socket Throughput
High performance on chip interconnect
Improved cache utilization
Integrated high speed memory
controllers
Increased Core Parallelism
Continue to improve parallelism
beyond SMT2
Improved execution unit utilization
4X system performance
in the same energy envelope
as the previous generation high-end system
0
500
1000
1500
2000
2500
3000
p6-595 p795
Relative rPerf Data
IBM Power Systems
9
eDRAM Technology
Enables POWER7 to provide 32MB of on-chip L3 Cache
IBMs eDRAM technology:
Improves on-processor memory performance
1/3 the space of conventional and 1/5 the standby power
of a conventional SRAM implementation
First-of-its-kind, on-chip memory technology
Fastest access time ever
IBM is effectively doubling microprocessor performance beyond what classical
scaling alone can achieve, said Dr. Subramanian Iyer, DE (Distinguished
Engineer)
eDRAM Cell
IBM Power Systems
10
Multi-threading Improvements with POWER7
Simultaneous Multi-threading
Intelligent thread design
Up to 4 hardware threads (SMT4)
with POWER7
Improved utilization per core
FX0
FX1
FP0
FP1
LS0
LS1
BRX
CRL
Single thread
FX0
FX1
FP0
FP1
LS0
LS1
BRX
CRL
POWER5 2 Way SMT
FX0
FX1
FP0
FP1
LS0
LS1
BRX
CRL
POWER7 4 Way SMT
10 Func Units
2 FX
2 FP
2 LS
1 BRX
1 CRL
1- DP
1 - VEX
12 Func Units
2 FX
4 FP
2 LS
1 BRX
1 CRL
1- DP
1 - VEX
Thread 1 Executing Thread 0 Executing No Thread Executing
Thread 3 Executing Thread 2 Executing
IBM Power Systems
11
0
0.5
1
1.5
2
ST SMT2 SMT4
+20-30%
Hyper-threading
POWER7 Multi-threading Options
Different applications have different
requirements for performance
POWER7 offers different SMT modes
to provide varying level of throughput
and thread performance
Dynamic Runtime SMT scheduling
Can Dynamically shift between modes
as required: ST/SMT2/SMT4
SMT on POWER7 is superior to
Hyper-threading (HT) on Intel x86
Standard Cache Option
All cores active
IBM Power Systems
12
POWER7 has Clear Performance Leadership on Major Workloads
(Per Socket vs. Best Published 4/18/10 Intel Offering)
0
100
200
300
400
500
600
700
N
o
r
m
a
l
i
z
e
d

U
n
i
t
s
SAP SD SpecIntRate SpecFPrate SpecJbb2005 TPCC
POWER7
Nehalem-EX
Nehalem-EP
Westmere-EP
Itanium
IBM Power Systems
13
Best SAP S&D 2-Tier Performance is on Power Systems
x86 SAP S&D 2-Tier Results
without virtualization
SAP USERS
40,000
30,000
20,000
10,000
0
HP DL580 G7
Nehalem-EX
4/32/64
Power 750
POWER
4/32/128
HP DL980 G7
Nehalem-EX
8/64/128
Power 780
POWER
8/64/256
4 sockets 8 sockets
Power 780
with DB2
4-socket
HP Nehalem-EX
#1
1.5X
Overall
#1
SAP users on SAP S&D 2-Tier
Systems are listed with processor chips/core/threads under system name; IBM Power System 780, 8p / 64c / 256t, POWER7, 3.8 GHz, 1024 GB memory, 37,000 SD users, dialog resp.: 0.98s, line items/hour: 4,043,670, Dialog steps/hour:
12,131,000, SAPS: 202,180, DB time (dialog/ update):0.013s / 0.031s, CPU utilization: 99%, OS: AIX 6.1, DB2 9.7, cert# 2010013; SUN M9000, 64p / 256-c / 512t, 1156 GB memory, 32,000 SD users, SPARC64 VII, 2.88 GHz, Solaris 10,
Oracle 10g , cert# 2009046; All results are 2-tier, SAP EHP 4 for SAP ERP 6.0 (Unicode) and valid as of 7/13/2010; Source: http://www.sap.com/solutions/benchmark/sd2tier.epx - See Power 780 benchmark details for more information
IBM Power Systems
14
Power Systems RAS provides the resiliency that x86 cant
Application/Partition RAS
No Yes Extended Error Handling
No Yes Survive Double Memory Failures
No Yes Selective Memory Mirroring
No
Yes Live Application Mobility
RAS Feature Power Systems Intel
Live Partition Mobility (vMotion) Yes Yes
Partition Availability Priority Yes No
System RAS
OS independent First Failure Data Capture Yes No
Memory Keys (including OS exploitation) Yes No
Processor RAS
Processor Instruction Retry Yes No
Alternate Processor Recovery Yes No
Dynamic Processor Deallocation Yes No
Dynamic Processor Sparing Yes No
Memory RAS
Chipkill Yes Yes
Redundant Memory Yes Yes
I/O RAS
I/O Adapter Isolation (PCI-Bus and TCEs) Yes No
See the following URLs for addition details:
http://www-03.ibm.com/systems/migratetoibm/systems/power/availability.html
http://www-03.ibm.com/systems/migratetoibm/systems/power/virtualization.html
IBM Power Systems
15
Intel Machine Check Architecture (MCA) Recovery
Shipped with Intel Xeon 7500 processor family
Detects CPU, memory and I/O errors
Requires the hypervisor to identify, correct and recover from otherwise fatal
system errors Future statement of support for VMware
VMware will be supporting Intels implementation of MCA recovery in future versions of VMware
vSphere to enhance consolidation of business critical workloads. When combined with VMware
vSphere, Intel's MCA will deliver new error recovering capabilities for improved reliability in large
memory systems helping customers accelerate their journey towards 100 percent virtualization to
achieve better efficiency, control and choice for their datacenters.
-- Dr. Stephen Herrod, CTO and SVP of R&D - May 2009
Currently provides additional information for a hardware machine check!
IBM Power Systems
16
Industry-Leading Energy Efficiency
POWER4 POWER5 POWER6 POWER7
P
e
r
f
o
r
m
a
n
c
e

p
e
r

W
a
t
t
POWER continues to accomplish major strides in energy efficiency
IBM Power Systems
17
Power Systems Advantages
POWER Hypervisor
Server Hardware
Balanced System
Design
Energy
Efficiency
High
Performance
Reliable
Hardware
Virtual Machines
IBM Power Systems
18 18
Hypervisor
IBM Power Systems
19
Industrial Strength Virtualization on Power Systems
The POWER Hypervisor is a thin bare metal hypervisor that virtualizes the system
processors and memory, and supports virtual networks within the server
One or more Virtual I/O Server partitions are used to virtualize the I/O adapters and
storage devices
POWER Hypervisor
Service
Processor
Processors
Memory
I/O Slots
Server
Hardware
Local Devices and Storage
Virtual Memory
Hardware
Management
Console
IBM i
Partitions
Unassigned
or
CoD
Resources
AIX
Partitions
Linux
Partitions
Virtual I/O
Servers
Virtual Adapters
Shared
Disks
Physical
Adapters
Virtual Processors
Virtual Networks
IBM Power Systems
20
PowerVM Provides Superior Availability
OS
Hypervisor
OS
Device
Driver
Proxy
PowerVM
Device
Driver
Proxy
Firmware (VIOS) provides I/O sharing
Hardware (TCEs) provide I/O isolation
Dual VIO Servers for redundancy
Message Passing
PCI-family adapters that
cannot be shared directly
TCEs (Translation Control Entries)
VMware
I/O virtualization and device drivers are part of
hypervisor reducing overall system availability
Failure of I/O adapter or device driver can
cause system outage or data corruption
OS OS OS OS
Hypervisor
Virtualizes I/O
Device
Driver
Device
Driver
Storage
Adapter
Network
Adapter
OS
Device
Driver
Proxy
Storage
Adapter
Network
Adapter
Virtual
I/O Server
Virtualize devices and adapters
Device
Drivers
Host
I/F
Device
Drivers
Host
I/F
Virtual
I/O Server
Hardware Isolation
IBM Power Systems
21
VMware is Improving Security and Reliability
ESXi is the embedded version of the ESX Server
Removed Linux-based service console
Equivalent functionality of the ESX Server
5% of the size of ESX Server
Loaded from USB or SD flash storage device (or local drive)
ESX 4.1 is the last release of the traditional ESX Server
Customers will have to migrate to ESXi 4.1 or later
IBM Power Systems
22
Security with the POWER Hypervisor
The POWER Hypervisor is stored and controlled by the Service Processor
The Hardware Management Console (HMC) is the only interface to the Service Processor
HMC Access to the Service Processor is controlled by a userid and password over SSL
Firmware updates are signed and downloaded from the IBM support website to the HMC
POWER Hypervisor
Service
Processor
Server Hardware
Hardware
Management
Console
OS OS OS
Virtual
I/O
Servers
Administrator
userid/password
Private
Network
Firmware Updates
IBM Power Systems
23
PowerVM Provides the Security Customers Demand for their
Mission-Critical Workloads
PowerVM has never had a single reported security vulnerability
0
20
40
60
80
100
120
N
u
m
b
e
r

o
f

r
e
p
o
r
t
e
d

s
e
c
u
r
i
t
y
v
u
l
n
e
r
a
b
i
l
i
t
i
e
s
VMware PowerVM
Source: National Vulnerability Database, http://nvd.nist.gov/
IBM Power Systems
24
How VMware Handles Hardware Errors
VMware is developed for commodity x86 hardware from AMD and Intel
Lack of integration with the hardware compared to Power Systems
VMware Kernel Panic
PSOD (Purple Screen of Death)
Decoding a Machine Check Exception (MCE) VMware KB 1005184
IBM Power Systems
25
Frequency of Critical Hypervisor Updates
ESXi400-200907001 08/06/2009
07/09/2009
09/24/2009
11/27/2009
01/05/2010
03/03/2010
04/01/2010
05/27/2010
06/10/2010
09/30/2010
ESXi400-200906001
ESXi400-200909001
ESXi-4.0.0-update01 EH350_038_038 10/30/09
ESXi400-200912001
ESXi400-201002001 EH350_049_038 03/10/10
ESXi400-201003001
ESXi400-201005001
upgrade-from-esxi4.0-4.0_update02 EH350_071_038 06/30/10
ESXi400-201009001
VMware ESXi 4 Updates
Power Systems Firmware
(updates to POWER Hypervisor)
Source: VMware Downloads for ESXi 4.0
IBM Power Systems
26
VMware Marketing Claim
Today, virtual machines on ESX 4 can scale to eight virtual CPUs, 256GB of
memory, and over 350,000 disk IOPS, while keeping overhead limited between
2 and 10 percent for the majority of applications.
SAP software also performs very well on VMware vSphere. Using the standard
SAP SD benchmark, a virtual machine running 1-4 virtual CPUs matches the
performance of physical servers with less than 10 percent overhead.
Virtualizing Business-Critical Applications on VMware vSphere Whitepaper
see http://www.vmware.com/files/pdf/VMW_10Q1_WP_vSPHERE_USLET_EN_R6_proof.pdf
IBM Power Systems
27
VMware Actual Performance with SAP S&D 2-Tier
31%
38%
Lower
Throughput
VMware
Server 4.0
1 VM
using 8 virtual
CPUs
(97%)
Fujitsu PRIMERGY Model RX300 S5, 2 processors
/ 8 cores / 16 threads, Intel Xeon Processor X5570,
2.93 GHz, 64 KB L1 cache and 256 L2 cache per
core, 8MB L3 cache per processor, 96 GB main
memory
2,056 2009029
CPU Utilization Central Server Configuration
SAP S&D
Benchmark
Users
SAP
Certification
Number
VMware
Server 4.0
1 VM
using 4 virtual
CPUs
(98%)
Fujitsu PRIMERGY Model RX300 S5, 2 processors
/ 8 cores / 16 threads, Intel Xeon Processor X5570,
2.93 GHz, 64 KB L1 cache and 256 L2 cache per
core, 8MB L3 cache per processor, 96 GB main
memory
1,144 2009028
99%
Fujitsu PRIMERGY Model TX300 S5/RX300 S5, 2
processors / 8 cores / 16 threads, Intel Xeon
Processor X5570, 2.93 GHz, 64 KB L1 cache and
256 L2 cache per core, 8MB L3 cache per
processor, 48 GB main memory
3,328 2009014
SAP S&D Standard Application Benchmark Results, Two-Tier Internet Configuration
see http://www.sap.com/solutions/benchmark/sd2tier.epx
IBM Power Systems
28
VMware has 40% lower throughput compared to PowerVM running the
same workload
VMware has 40% lower
throughput compared to
PowerVM on 8-way systems
running at 100% utilization,
running the same workload
and virtualized resources*
PowerVM runs workloads
more efficiently than VMware,
with superior resource
utilization
0
20000
40000
60000
80000
100000
120000
140000
160000
J
o
b
s
/
m
i
n
1 2 4 6 8
Number of CPUs
AIM7 Performance Benchmark
Single VM Scaling (Scale-up)
vSphere 4 on HP DL380 PowerVM on Power 750
HP DL380 G6
Power 750
* A Comparison of PowerVM and VMware Virtualization Performance, April 2010
http://www.ibm.com/systems/power/software/virtualization/whitepapers/compare_perf.html
IBM Power Systems
29
POWER Hypervisor Advantages
POWER Hypervisor
Server Hardware
Secure Integrated Stable
Virtual Machines
IBM Power Systems
30 30
Workloads
IBM Power Systems
31
AIX on Power Systems Minimizes Downtime
*Source: ITIC 2009 Global Server Hardware & Server OS Reliability Survey Results, July 7, 2009.
Complete details at http://itic-corp.com/blog/2009/07/itic-2009-global-server-hardware-server-os-reliability-survey-results/
Open Source Linux x86
HP UX 11/ HP Integrity
HP UX 11/ PA RISC
Sun Solaris / SPARC
IBM AIX POWER
Apple MAC
Red Hat Enterprise Linux x86
Windows Server 2008 x86
Windows Server 2003 x86
Downtime (Hours per Year)
IBM quality of service
99.997% uptime*
2.3X better than next UNIX
>10X better than x86-based platforms
We wanted to move away from
Windows onto a Unix platform that
would offer better security and stability
with less need for patching IBM was
the best possible choice.
Francisco Friderichsen Herrera, CTO
Agricola National
IBM Power Systems
32
VMware only provides Hot Add for CPU and Memory
Hot-add feature is only licensed for Advanced, Enterprise and Enterprise Plus
versions of vSphere
Not enabled by default at the VM level
Have to stop VM if it is already running
Hot-add for cpu and memory is dependent on the Guest O/S
May still require a reboot of the guest O/S to recognize the new cpu or memory
Not compatible with VMware Fault Tolerance
No Hot Remove of CPU and Memory if the load is reduced!
With PowerVM, CPU and Memory can be added dynamically to a running
workload
IBM Power Systems
33
VMotion Only Improves Availability for Planned Outages
VMotion is useful for
Transferring workloads for hardware maintenance
Rebalancing workloads within a cluster
Refreshing hardware for an existing cluster
Limitations for VMware
CPUs from a single vendor , either Intel or AMD
Fairly new CPUs with hardware virtualization support (AMD-V or Intel VT) enabled
Enhanced VMotion Compatibility (EVC) support for AMD-V Extended Migration or
Intel FlexMigration for VMotion between generations of processors or create a
processor baseline by masking specific capabilities
PowerVM is less problematic to move between generations of POWER
(POWER6, POWER7 and future) processors
IBM Power Systems
34
VMware Does Not Provide High Availability
VMware High Availability restarts all the virtual machines on another host if an
ESX host becomes unavailable
VMware High Availability is positioned as an alternative for high availability
clustering solutions
If you use VMware HA, be aware that:
VMware HA handles ESX Server host hardware failure and does not monitor
the status of the SAP services such as the SAP enqueue service or the
database; these must be monitored by separate third party clustering software.
Virtual machines are automatically restarted on the failover ESX Server hosts
but the SAP application on the virtual machine is not. Separate startup scripts or
services in the guest operating system can enable automatic start up of the SAP
applications.
Best Practice Guidelines for SAP Solutions on VMware Infrastructure Whitepaper p25
http://www.vmware.com/files/pdf/whitepaper_SAP_bestpractice_jan08.pdf
IBM Power Systems
35
PowerVM Supports the Most Demanding Growth Requirements
Workloads on PowerVM can scale
dynamically from a minimum of
1/10 CPU to our largest server
with 256 CPUs to support the most
demanding growth requirements
Workloads are limited by the
available system resources not
limits in technology. The resources
can be added without interruption
using unallocated or CoD
resources
0
50
100
150
200
250
300
P
h
y
s
i
c
a
l

C
P
U
s
Workload Scalability
2
5
6
0
x
IBM Power Systems
36
PowerVM Does Not Limit the Size of Virtualized Workloads
A virtual machine in VMware
Enterprise Plus Edition is
limited to 8 virtual CPUs.
A virtual machine is limited
to 8 cores without Hyper-
threading or only 4 cores
with Hyper-threading
PowerVM is only limited by
the resources in the system
0
100000
200000
300000
400000
500000
600000
J
o
b
s
/
m
i
n
1 8 16 24 32
Number of CPUs
AIM7 Performance Benchmark
Single VM Scaling (Scale-up)
vSphere 4 on HP DL380 PowerVM on Power 750
PowerVM scales linearly
* A Comparison of PowerVM and VMware Virtualization Performance, April 2010
http://www.ibm.com/systems/power/software/virtualization/whitepapers/compare_perf.html
VMware limit
is 8 vCPUs
IBM Power Systems
37
Workload Advantages
POWER Hypervisor
Server Hardware
Reliable
O/S
Dynamic
Workloads
Higher
Consolidation
Ratio
Workloads
Scale
Vertically
Workload
Growth
Virtual Machines
IBM Power Systems
38 38
Customer Examples
IBM Power Systems
39
Rotkappchen selects Power Systems and DB2
to grow their SAP landscape
With the IBM Power architecture, we
only need two physical servers to run
the entire environment. This makes the
solution highly energy-efficient,
compared to a distributed x86-based
architecture which is not only an
advantage in terms of operational costs,
but also helps to reduce the carbon
footprint
Andre Birrenbach, CIO
Rotkappchen-Mumm Sektkellereien
Challenge
An Rotkappchen acquisition created
challenges of integrating the new
companys systems and processes into
their infrastructure to reduce operational
costs and increase standardization and
efficiency.
Solution
Due to the excellent performance and
stability they deployed the production
instances of SAP test, development, quality
assurance and legacy applications on IBM
Power Systems i550 with four POWER6
processors and have retired all the servers
from the acquisition.
IBM Power Systems
40
Inventia reduces IT costs and increases efficiency by moving from
Windows to Power Systems and DB2
Challenge
Outpacing capacity of existing management
systems due to rapid annual growth.
Application maintenance and support costs
were rising with an infrastructure of hardware
running Windows with fragmented data
structures
Solution
Replaced a fragmented collection of systems
and databases with the SAP for
Pharmaceuticals running on a single
integrated instance on an IBM Power System
with DB2. Data is immediately available for
analysis throughout the enterprise and audit
preparation requires 60% less effort
The IBM Power Systems server is
capable of delivering reports with no
interruption to normal production. The
result is lower business costs, improved
information, and better business
planning
Dr. Ankur Shah, Executive Director
Inventia Healthcare
IBM Power Systems
41 41
TCA/TCO
Calculations for SAP
IBM Power Systems
42
Assumptions for TCA/TCO Calculations for SAP
Comparison of potential SAP deployment options
Windows/x86, Windows/VMware/x86 and AIX/PowerVM/Power 740
SAP Licenses are not included as the software licenses are based on the
number of users
Database costs are not included as SAP is an OEM for SQL Server/DB2/Oracle
Storage costs are not included as the costs are assumed the same for all
solutions
Administration costs are not included in the TCO calculations as any staff
savings would be deployed on other projects
VMware does reduce the physical number of x86 servers but increases the
software costs for TCA/TCO
SAP overhead is greater than 2-10% for VMware on x86 based on benchmark
results
IBM Power Systems
43
Totals $854,329 $394,734 $180,718
$0
$100,000
$200,000
$300,000
$400,000
$500,000
$600,000
$700,000
$800,000
$900,000
1 year TCA
HP DL580 G7 vs. Power 740
2500 OS Cl i ent Li censes
Vi rtual i zati on Mai ntenance
Vi rtual i zati on
OS Mai ntenance
OS
HWMA
HW
2500 OS Client Licenses $150,000 $150,000 $0
Virtualization Maintenance $0 $5,456 $2,240
Virtualization $0 $23,236 $13,600
OS Maintenance $45,865 $11,466 $4,000
OS $114,312 $45,865 $16,000
HWMA $0 $0 $0
HW $544,152 $158,711 $144,878
DL 580 G7 + Windows
DL 580 G7 + VMware +
Windows
Power 740 Express +
AIX
IBM Power Systems
44
One year TCA savings of $214,016 when deploying SAP on two, 16-core IBM
Power 740 Express systems utilized at 90% with PowerVM on AIX vs. seven, 16-core HP
DL580 G7 utilized at 70% with VMware on Microsoft Windows Server 2008.
Power 740 Express HP ProLiant DL580 G7
$50k less software license / maintenance costs
Power Systems lowers clients total cost of acquisition by 54%
IBM Power Systems
45
Three year TCO savings of $272,294 when deploying SAP on two, 16-core IBM
Power 740 Express systems utilized at 90% with PowerVM on AIX vs. seven, 16-core HP
DL580 G7 utilized at 70% with VMware on Microsoft Windows Server 2008.
Power 740 Express HP ProLiant DL580 G7
$72k less software license / maintenance costs
$17k less energy costs
71% less space
Power Systems lowers clients total cost of ownership by 57%
IBM Power Systems
46
Improved TCA/TCO for SAP Deployments using PowerVM
Higher utilization of Power Systems reduces the number of physical systems for
the same number of SAP users with significant reductions in space and energy
costs
SAP License charges are the same for equal numbers of users but database
charges for DB2 and Oracle can be reduced with less hardware as the software
is licensed per core
The virtualized workloads can grow seamlessly to match demand rather then
limited by virtualization technology
It is not uncommon for PowerVM customers to combine different workloads as
well as production and non-production workloads on the same system, making
the best use of the available resources
IBM Power Systems
47 47
Summary
IBM Power Systems
48
PowerVM Advantages
POWER Hypervisor
Server Hardware
Secure Integrated
High
Performance
Energy
Efficiency
Balanced
System Design
Stable
Reliable
O/S
Dynamic
Workloads
Reliable
Hardware
Higher
Consolidation
Ratio
Workloads
Scale
Vertically
Workload
Growth
Virtual Machines
IBM Power Systems
49
Summary
Hardware, firmware and O/S are developed, tested and supported together
Virtualization is always on with Power Systems
Individual workloads scale beyond 4 or 8 physical processors on PowerVM
PowerVM allows workloads to dynamically add and remove resources to meet
changing business requirements
Decreased unplanned downtime due to reliable hardware, hypervisor and
operating systems
VMware 4.0 overhead is greater than 2-10% for critical business applications
70+% attach rate for PowerVM on all new POWER7 systems
Higher utilization with PowerVM means fewer systems and a simpler
environment to manage
IBM Power Systems
50
Backup
IBM Power Systems
51
16 cores, 2.27 GHz
128 GB memory
70% Utilization
36% VMware Overhead
Seven HP ProLiant DL580 G7
97%
64%
Utilization
Efficiency
AIX 7.1
Windows Server
2008
Operating System Company /
System
Virtualization Software Rack
Units
Watts
HP ProLiant
DL580 G7
1
VMware vSphere 4.1 Enterprise,
VMware vCenter Server 4
Standard
4 609
IBM Power 740
Express
1
PowerVM Express Edition, IBM
Systems Director Express Edition
4 653
More performance per core means fewer cores and lower software costs
16 cores, 3.55 GHz
256 GB memory
90% Utilization
Two Power 740 Express
1. Storage not included
2. Performance for both Power 740 and HP DL580 G7 are estimated using published SAP benchmarks for Power 750 and HP ProLiant DL580 G7 systems.
IBM Power Systems
52
Totals $1,128,467 $481,781 $209,487
$0
$200,000
$400,000
$600,000
$800,000
$1,000,000
$1,200,000
3 year TCO
HP DL580 G7 vs. Power 740
2500 OS Cl i ent Li censes
Faci l i ti es
Vi rtual i zati on Mai ntenance
Vi rtual i zati on
OS Mai ntenance
OS
HWMA
HW
2500 OS Client Licenses $150,000 $150,000 $0
Facilities $182,408 $53,202 $16,289
Virtualization Maintenance $0 $16,368 $6,720
Virtualization $0 $23,236 $13,600
OS Maintenance $137,595 $34,399 $12,000
OS $114,312 $45,865 $16,000
HWMA $0 $0 $0
HW $544,152 $158,711 $144,878
DL 580 G7 +
Windows
DL 580 G7 + VMware
+ Windows
Power 740 Express +
AIX
IBM Power Systems
53
This document was developed for IBM offerings in the United States as of the date of publication. IBM may not make these offerings available in
other countries, and the information is subject to change without notice. Consult your local IBM business contact for information on the IBM
offerings available in your area.
Information in this document concerning non-IBM products was obtained from the suppliers of these products or other public sources. Questions
on the capabilities of non-IBM products should be addressed to the suppliers of those products.
IBM may have patents or pending patent applications covering subject matter in this document. The furnishing of this document does not give
you any license to these patents. Send license inquires, in writing, to IBM Director of Licensing, IBM Corporation, New Castle Drive, Armonk, NY
10504-1785 USA.
All statements regarding IBM future direction and intent are subject to change or withdrawal without notice, and represent goals and objectives
only.
The information contained in this document has not been submitted to any formal IBM test and is provided "AS IS" with no warranties or
guarantees either expressed or implied.
All examples cited or described in this document are presented as illustrations of the manner in which some IBM products can be used and the
results that may be achieved. Actual environmental costs and performance characteristics will vary depending on individual client configurations
and conditions.
IBM Global Financing offerings are provided through IBM Credit Corporation in the United States and other IBM subsidiaries and divisions
worldwide to qualified commercial and government clients. Rates are based on a client's credit rating, financing terms, offering type, equipment
type and options, and may vary by country. Other restrictions may apply. Rates and offerings are subject to change, extension or withdrawal
without notice.
IBM is not responsible for printing errors in this document that result in pricing or information inaccuracies.
All prices shown are IBM's United States suggested list prices and are subject to change without notice; reseller prices may vary.
IBM hardware products are manufactured from new parts, or new and serviceable used parts. Regardless, our warranty terms apply.
Any performance data contained in this document was determined in a controlled environment. Actual results may vary significantly and are
dependent on many factors including system hardware configuration and software design and configuration. Some measurements quoted in this
document may have been made on development-level systems. There is no guarantee these measurements will be the same on generally-
available systems. Some measurements quoted in this document may have been estimated through extrapolation. Users of this document
should verify the applicable data for their specific environment.
Revised September 26, 2006
Special notices
IBM Power Systems
54
IBM, the IBM logo, ibm.com AIX, AIX (logo), AIX 6 (logo), AS/400, BladeCenter, Blue Gene, ClusterProven, DB2, ESCON, i5/OS, i5/OS (logo), IBM Business Partner
(logo), IntelliStation, LoadLeveler, Lotus, Lotus Notes, Notes, Operating System/400, OS/400, PartnerLink, PartnerWorld, PowerPC, pSeries, Rational, RISC
System/6000, RS/6000, THINK, Tivoli, Tivoli (logo), Tivoli Management Environment, WebSphere, xSeries, z/OS, zSeries, AIX 5L, Chiphopper, Chipkill, Cloudscape, DB2
Universal Database, DS4000, DS6000, DS8000, EnergyScale, Enterprise Workload Manager, General Purpose File System, , GPFS, HACMP, HACMP/6000, HASM, IBM
Systems Director Active Energy Manager, iSeries, Micro-Partitioning, POWER, PowerExecutive, PowerVM, PowerVM (logo), PowerHA, Power Architecture, Power
Everywhere, Power Family, POWER Hypervisor, Power Systems, Power Systems (logo), Power Systems Software, Power Systems Software (logo), POWER2,
POWER3, POWER4, POWER4+, POWER5, POWER5+, POWER6, System i, System p, System p5, System Storage, System z, Tivoli Enterprise, TME 10, Workload
Partitions Manager and X-Architecture are trademarks or registered trademarks of International Business Machines Corporation in the United States, other countries, or
both. If these and other IBM trademarked terms are marked on their first occurrence in this information with a trademark symbol ( or ), these symbols indicate U.S.
registered or common law trademarks owned by IBM at the time this information was published. Such trademarks may also be registered or common law trademarks in
other countries. A current list of IBM trademarks is available on the Web at "Copyright and trademark information" at www.ibm.com/legal/copytrade.shtml
The Power Architecture and Power.org wordmarks and the Power and Power.org logos and related marks are trademarks and service marks licensed by Power.org.
UNIX is a registered trademark of The Open Group in the United States, other countries or both.
Linux is a registered trademark of Linus Torvalds in the United States, other countries or both.
Microsoft, Windows and the Windows logo are registered trademarks of Microsoft Corporation in the United States, other countries or both.
Intel, Itanium, Pentium are registered trademarks and Xeon is a trademark of Intel Corporation or its subsidiaries in the United States, other countries or both.
AMD Opteron is a trademark of Advanced Micro Devices, Inc.
Java and all Java-based trademarks and logos are trademarks of Sun Microsystems, Inc. in the United States, other countries or both.
TPC-C and TPC-H are trademarks of the Transaction Performance Processing Council (TPPC).
SPECint, SPECfp, SPECjbb, SPECweb, SPECjAppServer, SPEC OMP, SPECviewperf, SPECapc, SPEChpc, SPECjvm, SPECmail, SPECimap and SPECsfs are
trademarks of the Standard Performance Evaluation Corp (SPEC).
NetBench is a registered trademark of Ziff Davis Media in the United States, other countries or both.
AltiVec is a trademark of Freescale Semiconductor, Inc.
Cell Broadband Engine is a trademark of Sony Computer Entertainment Inc.
InfiniBand, InfiniBand Trade Association and the InfiniBand design marks are trademarks and/or service marks of the InfiniBand Trade Association.
Other company, product and service names may be trademarks or service marks of others.
Revised April 24, 2008
Special notices (cont.)
IBM Power Systems
55
The IBM benchmarks results shown herein were derived using particular, well configured, development-level and generally-available computer systems. Buyers should
consult other sources of information to evaluate the performance of systems they are considering buying and should consider conducting application oriented testing. For
additional information about the benchmarks, values and systems tested, contact your local IBM office or IBM authorized reseller or access the Web site of the benchmark
consortium or benchmark vendor.
IBM benchmark results can be found in the IBM Power Systems Performance Report at http://www.ibm.com/systems/p/hardware/system_perf.html .
All performance measurements were made with AIX or AIX 5L operating systems unless otherwise indicated to have used Linux. For new and upgraded systems, AIX
Version 4.3, AIX 5L or AIX 6 were used. All other systems used previous versions of AIX. The SPEC CPU2006, SPEC2000, LINPACK, and Technical Computing
benchmarks were compiled using IBM's high performance C, C++, and FORTRAN compilers for AIX 5L and Linux. For new and upgraded systems, the latest versions of
these compilers were used: XL C Enterprise Edition V7.0 for AIX, XL C/C++ Enterprise Edition V7.0 for AIX, XL FORTRAN Enterprise Edition V9.1 for AIX, XL C/C++
Advanced Edition V7.0 for Linux, and XL FORTRAN Advanced Edition V9.1 for Linux. The SPEC CPU95 (retired in 2000) tests used preprocessors, KAP 3.2 for FORTRAN
and KAP/C 1.4.2 from Kuck & Associates and VAST-2 v4.01X8 from Pacific-Sierra Research. The preprocessors were purchased separately from these vendors. Other
software packages like IBM ESSL for AIX, MASS for AIX and Kazushige Gotos BLAS Library for Linux were also used in some benchmarks.
For a definition/explanation of each benchmark and the full list of detailed results, visit the Web site of the benchmark consortium or benchmark vendor.
TPC http://www.tpc.org
SPEC http://www.spec.org
LINPACK http://www.netlib.org/benchmark/performance.pdf
Pro/E http://www.proe.com
GPC http://www.spec.org/gpc
VolanoMark http://www.volano.com
STREAM http://www.cs.virginia.edu/stream/
SAP http://www.sap.com/benchmark/
Oracle Applications http://www.oracle.com/apps_benchmark/
PeopleSoft - To get information on PeopleSoft benchmarks, contact PeopleSoft directly
Siebel http://www.siebel.com/crm/performance_benchmark/index.shtm
Baan http://www.ssaglobal.com
Fluent http://www.fluent.com/software/fluent/index.htm
TOP500 Supercomputers http://www.top500.org/
Ideas International http://www.ideasinternational.com/benchmark/bench.html
Storage Performance Council http://www.storageperformance.org/results
Revised March 12, 2009
Notes on benchmarks and values
IBM Power Systems
56
Revised April 2, 2007
Notes on performance estimates
rPerf for AIX
rPerf (Relative Performance) is an estimate of commercial processing performance relative to other IBM UNIX systems. It is derived from an
IBM analytical model which uses characteristics from IBM internal workloads, TPC and SPEC benchmarks. The rPerf model is not
intended to represent any specific public benchmark results and should not be reasonably used in that way. The model simulates some of
the system operations such as CPU, cache and memory. However, the model does not simulate disk or network I/O operations.
rPerf estimates are calculated based on systems with the latest levels of AIX and other pertinent software at the time of system
announcement. Actual performance will vary based on application and configuration specifics. The IBM eServer pSeries 640 is the
baseline reference system and has a value of 1.0. Although rPerf may be used to approximate relative IBM UNIX commercial processing
performance, actual system performance may vary and is dependent upon many factors including system hardware configuration and
software design and configuration. Note that the rPerf methodology used for the POWER6 systems is identical to that used for the
POWER5 systems. Variations in incremental system performance may be observed in commercial workloads due to changes in the
underlying system architecture.
All performance estimates are provided "AS IS" and no warranties or guarantees are expressed or implied by IBM. Buyers should consult
other sources of information, including system benchmarks, and application sizing guides to evaluate the performance of a system they are
considering buying. For additional information about rPerf, contact your local IBM office or IBM authorized reseller.

VM24

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

VM24

Uploaded by

Copyright:

Available Formats

2010 IBM Corporation

IBM Power Systems

You might also like