Add Memory, Improve Performance, and Lower Costs with IBM MAX5 Technology

© Copyright IBM Corp. 2012. All rights reserved. ibm.com/redbooks 1
Redpaper
Add Memory, Improve Performance, and
Lower Costs with IBM MAX5 Technology
Executive summary
The need for computer memory is increasing at a tremendous rate as databases continue to
grow. With the increased use of virtualization, applications become memory-constrained and
performance bottlenecks occur. At the same time, x86 processors are evolving with more
cores and more thread capability. Without adequate memory, these processors are
underutilized. This lack of memory is occurring more frequently and with more intensity as
applications, databases, and technologies grow and change.
A previous solution was to purchase additional computers, and then distribute applications
among them to use the combined memory capacity. However, the hardware, data center
space, power and cooling requirements, and licensing were costly, and still the computers
were underutilized.
IBM® is solving this issue with a new server technology in its IBM System x® eX5 class of
server: the MAX5 memory expansion unit. This unit is an IBM-only, System x technology that
adds memory apart from the processor. With no added processors, MAX5 is designed to
increase productivity and reduce costs in a way that other servers of this class cannot.
At the IBM Strategy and Technology Lab (STL), we wanted to test how MAX5 improved
system performance on web-based and database workloads. We used a production-quality
IBM System x3690 X5 server, and we tested with and without MAX5. The results presented in
this IBM Redpaper™ publication are from actual measurements taken during our STL testing.
Tim Bohn
Dr. Howard Operowsky
Karen Lawrence

2 Add Memory, Improve Performance, and Lower Costs with IBM MAX5 Technology
In this paper, we demonstrate how the additional memory afforded by MAX5 improved the
performance and costs of database and web-facing applications in our tests. In particular, our
results show the following benefits:
򐂰 Response time for business intelligence reports was 1.5 - 2.8 times faster
򐂰 Cost of business intelligence reports was decreased by 31% - 64%
򐂰 Throughput of web-facing applications was 2.4 - 4.9 times greater
򐂰 Read/write response time was decreased by 60% - 80%
򐂰 It can cost 49% more for a web-facing application system without MAX5 to match the
performance of a system with MAX5
This publication is directed to IT professionals and decision-makers, such as CEOs, CIOs,
CFOs, information architects, business intelligence administrators, IBM DB2® database
administrators, IBM clients, and IBM Business Partners.
How MAX5 technology works
IBM System x MAX5 technology1
is based on eX5, the next generation of Enterprise IBM
X-Architecture®. The design of the MAX5 increases the number of dual inline memory
modules (DIMMs) supported in the system up to 32. With MAX5 and 32 GB DIMMs, you can
add as much as 1 TB of memory to an existing eX5 computer to support large,
memory-constrained applications. This industry-first technology separates memory from the
processor by using the following features:
򐂰 The advanced eX5 chipset:
– The design of the MAX5 increases the number of DIMMs supported in the system up
to 32. With MAX5 and 32 GB DIMMs, you can add as much as 1 TB of memory to an
existing eX5 computer to support large, memory-constrained applications.
– Attaches to QuickPath Interconnect (QPI) links as a node controller, thus giving the
chipset direct access to all CPU bus transactions.
– Maintains a directory of the last-level cache of each central processing unit (CPU), thus
allowing the chipset to respond to memory requests before the end of a broadcast
snoop cycle, improving performance.
򐂰 An additional 16 channels of memory bandwidth
These added channels provide a significant boost to overall throughput and additional
memory performance. Even on a system that is not memory-constrained, the additional
channels allow the system to reach higher memory capacity by using smaller, lower-priced
chips. The result is a more economical system.
As this paper shows, these benefits translate into faster response times, faster throughput,
and lower costs.
1
IBM eX5 Portfolio Overview: IBM System x3850 X5, x3950 X5, x3690 X5, and BladeCenter HX5, REDP-4650:
http://www.redbooks.ibm.com/abstracts/redp4650.html

Add Memory, Improve Performance, and Lower Costs with IBM MAX5 Technology 3
The System x3690 X5 with the MAX5 attached is shown in Figure 1.
Figure 1 IBM System x3690 X5 with MAX5 memory expansion unit attached beneath it
About our products
Before we describe our tests, we look at some of the hardware and software components we
used in the tests and describe why we chose them.
IBM System x3690 X5 server
We used System x3690 X5 servers because of their inherent power. These servers are
two-socket, rack-mount servers with up to 10 core Intel Xeon EX processors. Some x3690 X5
server models are combined with MAX5 (with its 32 GB of RAM), to supply up to 1 TB of
memory.
This server has up to 16 hard disk drives (HDDs), supporting 6.35 cm (2.5 in.), hot-swap,
internal Serial Attached SCSI/Serial Advanced Technology Attachment (SAS/SATA) drives,
totaling up to 8 TB. The server also provides up to 24 solid-state drives (SSDs), supporting
4.57 cm (1.8 in.), hot-swap, solid-state drives, by using up to three IBM eXFlash storage
options2
. Each solid-state drive is designed to sustain approximately 4,300 input/output
operations per second (IOPS). This IOPS rate is more than 10 times the rate achieved by an
enterprise 15,000 RPM, 6.35 cm (2.5 in.), 146.8 GB SAS HDDs (420 IOPS).
IBM MAX5 for System x
MAX5 for System x is a 1U memory expansion drawer. It adds 32 DIMM slots, the advanced
eX5 chipset, and 16 additional channels of memory bandwidth, as explained in “How MAX5
technology works” on page 2. MAX5 is an optional unit for the IBM eX5 line of servers. It is
also available as an option on all models of the System x3690 X5.
2
IBM Unveils Industry’s First Systems that Rewrite Economics of ‘Industry-Standard’ Computing, March 2, 2010:
http://www.ibm.com/press/us/en/pressrelease/29570.wss

MAX5 for System x is scalable. In an IT environment, the base System x3690 X5 (without
MAX5) with 16 DIMMs can serve as a starting point. When your IT needs expand, you can
install an additional 16 DIMMs in your System x3690 X5. As your IT needs expand further,
you can install the 1U MAX5 memory expansion unit, thus adding 32 DIMM sockets. The total
number of sockets available to the system is then 64 DIMMs in only 3U of rack space. Using
16 GB DIMMs, your system has 1 TB of RAM.3
DB2
DB2 is a family of IBM relational database management system (RDBMS) products known for
a number of factors:
򐂰 The industry-leading database performance of DB2 is a significant reason for its use in
analytical and online transaction processing (OLTP) workloads. Such performance is a
factor in overall system performance. DB2 is also appropriate for mixed workloads on
distributed systems.
򐂰 DB2 is scalable and as equally effective for single-user applications as it is in large,
multiuser environments. Users can run structured query language (SQL) queries
simultaneously on DB2.
򐂰 DB2 is efficient. With its application development and management tools, you can
streamline database administrator tasks, detect emergent problems, offload inactive data,
and improve the speed of backup and recovery processes. This efficiency translates to
lower operational costs.
򐂰 DB2 supports multiple operating systems, including Microsoft Windows, Linux, UNIX, and
IBM z/OS®. Administration tasks can be carried out at a command line or the GUI, which
is a multi-platform Java client with multiple wizards for further increasing efficiency.
IBM WebSphere eXtreme Scale
WebSphere eXtreme Scale is a low-latency, elastic, in-memory data grid that, to an
application, looks like single, virtualized, in-memory cache. We chose IBM WebSphere®
eXtreme Scale for our test because it takes full advantage of the additional memory provided
by MAX5. A WebSphere eXtreme Scale grid is used effectively as a data cache for a
database or other data source, which is generally slower to respond to requests due to the
need to access data on a hard disk.
The elastic scalability of WebSphere eXtreme Scale is possible by using distributed object
caching. The grid is implemented by combining the heap spaces of a collection of Java virtual
machines (JVMs), which WebSphere eXtreme Scale calls containers. These containers can
be located in the same physical machine, or they can be dispersed geographically among
different physical machines. The data objects in the grid are partitioned among the heaps of
the JVMs. Applications are not affected by which JVM contains the data because the grid is
presented as a single entity. The grid can grow or shrink by adding or removing JVMs. Data in
the grid can be replicated to enhance reliability.
In an OLTP setting, WebSphere eXtreme Scale can process billions of transactions a day. It
does so by managing data and business logic across multiple servers and virtualized
environments. WebSphere eXtreme Scale also delivers predictable response time.4
3 IBM System x3690 X5 at-a-glance guide, January, 2012:
http://www.redbooks.ibm.com/abstracts/tips0818.html
4
WebSphere eXtreme Scale Best Practices for Operation and Management, SG24-7964:
http://www.redbooks.ibm.com/abstracts/sg247964.html?Open

About our tests
Two series of tests were conducted with and without MAX5: one tested database system
performance; the other tested web-facing transaction performance. This section describes
the overall purposes of the tests and their setups, configurations, procedures, and results.
Tests for database system performance
In this test, our goal was to determine the impact of MAX5 on performance. This test followed
a serial execution test in which successive reports were generated. The cost and
performance of the system with and without MAX5 also were compared.
For these tests, we used a production-quality System x3690 X5 server and IBM DB2 to
determine the effects of MAX5 on a 1.3 TB data warehouse in a business analytics
environment.
A simplified view of a System x3690 X5 connected to a MAX5 is shown in Figure 2.
Figure 2 A System x3690 X5 connected to a MAX5
Test setup and configuration
The server used in the test was an IBM System x3690 X5, with two, eight-core, Intel Xeon
X7560 processors, for a total of 16 cores. Two memory configurations were used:
򐂰 With MAX5: 512 GB RAM internal to the System x3690 X5, plus 512 GB RAM installed in
the MAX5 and attached to the System x3690 X5, for a total of 1 TB RAM
򐂰 Without MAX5: 512 GB RAM internal to the System x3690 X5
Memory
Buffer
Memory
Buffer
Memory
Buffer
Memory
Buffer
Memory
Buffer
Memory
Buffer
Memory
Buffer
Memory
Buffer
MAX5
MemoryExpansion
Memory
Buffer
Memory
Buffer
Memory
Buffer
Memory
Buffer
CPU
Memory
Buffer
Memory
Buffer
Memory
Buffer
Memory
Buffer
CPU
QPI
QPI QPI

We used HDDs on the server for the operating system, applications, database logs, and
database temp space. An IBM storage controller with 32 HDDs was attached for the
database. This configuration provided plenty of storage I/O throughput. The database was a
1.3 TB data warehouse that used IBM DB2 V9.7.4 and ran Red Hat Enterprise Linux V5.6.
For the driver, a client machine played the role of a user that sent report requests in the form
of sets of SQL queries. It was used to send the report requests and record the amount of time
taken for the database server to complete the instruction. The hardware and software
configuration used in our testing is shown in Figure 3.
Figure 3 Test hardware and software configuration
Business Intelligence Day
We created an environment that simulated a business intelligence (BI) workload, called BI
Day, which consisted of SQL queries from IBM Cognos® reports. These SQL queries are
submitted by Cognos to the database management system (DBMS) to generate a report.
Each report consisted of one to five SQL queries. The reports were designed to fall into these
typical reporting scenarios:
򐂰 Simple: Minimal, fast-running reports, dashboards, and ad hoc reports. The reports take
less than one second to run.
򐂰 Intermediate: Advanced reports that require predicate (a where clause) evaluation over a
large fact table, then join an aggregation of a relatively small result set (only a small
portion of the records are retrieved). These reports take from one second to one minute to
run.
򐂰 Complex: Expert and resource-intensive reports, requiring multiple joins and aggregations
on the full fact table (every row must be retrieved and operated on). These reports take
15 - 60 minutes to run.
We ran the first series of tests with the MAX5 attached and 1 TB total RAM. With this
configuration, nearly all of the data warehouse (about 97%) was buffered in memory.
We ran the second series of tests without MAX5 and used the standard memory configuration
of 512 GB. Here, more data than can be buffered in memory was present. The result was disk
I/O whenever the data was not present in memory (about 20% of the time) and needed to be
retrieved from disk. Both test series were run over the course of one week.
Client – IBM System x3250 M3
Client Driver
STL BI Day workload
IBM System x3690 X5
Intel Xeon Processor X7560
2.27 GHz, 2 sockets, 16 cores
512 GB RAM + 512 GB RAM (MAX5)
1 GB
IBM DB2 9.7FP4 (64 bit)
Red Hat Enterprise Linux 5.6
Test(s) disk configuration:
- OS/Apps: 1 HDD
- DB: DS5100 (32 HDDs)
- DB Logs: HDD
- DB Temp: HDD

Shown in Figure 4, the serial execution test simulates a user who executes each of these
reports consecutively and records the actual response time.
Figure 4 The serial execution test measured report performance for individual reports
Results
The test results demonstrated that MAX5 can cache up to twice as much information in
memory, thus reducing disk I/O and increasing performance and productivity. Having
available extra memory for a business analytics and data warehouse workload was valuable.
Response time was improved by caching the data in memory rather than continually reading
from disk I/O.
With MAX5 for business intelligence workloads, you can produce twice as many reports, as
we showed in our test over a one-week period (by using actual results). Therefore, over a
three-year period, the cost of reports can be 31% - 64% lower with the addition of MAX5 (by
using extrapolated results).
executes
intermediate reports
Intermediate
Report
9
Intermediate
Report
10
Intermediate
Report
11
1 user executes
complex reports
Complex
Report
1
Complex
Report
3
1 User
executes
simple reports
Simple
Report
6
Simple
Report
2
Simple
Report
4
Simple
Report
5
Single
Connection
Each report
executes one
or more queries
…then...
…then…
„ Reports generated by Cognos
„ Single user, serial execution test
„ Each query runs serially (each is
completed before the next query
begins)
Data Server

MAX5 turns reports around faster
As shown in Figure 5, with MAX5 attached, and when the response times for the three report
types are averaged, reports were completed 1.5 - 2.8 times faster. Of course, your
performance numbers might vary based on workload profiles.
Figure 5 Comparison of report response times: MAX5 (1 TB) versus standard memory (512 GB)
What this means to you as the company that needs to produce reports is clear: in a one-week
period, an organization can produce 432 complex reports with MAX5, but only 290 such
reports without MAX5. Add intermediate reports and many simple reports and you can see
that the company can generate twice as many reports in one week with MAX5. Without
MAX5, your organization needs another complete server setup, including software licenses
for additional cores, and data center space. Additional cooling and power needs also increase
operating costs.
Suppose you have the hardware and software we used in this test and you need to expand.
You can choose either option to meet your needs:
򐂰 Purchase an additional server and the software, software licensing, maintenance, data
center space, cooling, and power for the server.
򐂰 Purchase MAX5 and the extra 512 GB RAM. With this option, you do not incur expenses
for an additional server (hardware, software, software licensing, maintenance, data center
space, cooling, and power). Rather, you incur expenses for the MAX5 drawer and 512 GB
RAM only.
The difference in the results of the options is dramatic: the first option doubles your cost, a
100% increase; the second option increases your expenses by only 3%.
Average Complex Report
Performance
Serial Execution Test (1.3 TB)
(Lower is Better)
1.5x Faster
Average Intermediate Report
Performance
(Lower is Better)
1.7x Faster
33.41
19.29
0
5
10
15
20
25
30
35
40
Response
Time
(seconds)
Std Memory (512 GB) MAX5 (1 TB)
Average Simple Report
Performance
(Lower is Better)
2.8x Faster
0.56
0.2
0
0.1
0.2
0.3
0.4
0.5
0.6
Response
Time
(seconds)
2088
1400
0
500
1000
1500
2000
2500
Response
Time
(seconds)

For more information about the cost savings achieved when using MAX5, see “MAX5 reduces
the total cost of acquisition” on page 12.
MAX5 lowers the cost of reports
Although MAX5 decreases the time needed to generate reports, the question remains
whether it is worth the investment. Consider the total cost of hardware and software over a
three-year period, and the total number of reports that can be generated. The cost of reports
is 31% - 64% lower for the MAX5-outfitted server. The difference in the cost of reports over
that three-year period is shown in Figure 6. Your performance numbers might vary based on
workload profiles.
Figure 6 Comparison of cost of reports: MAX5 (1 TB) versus standard memory (512 GB)
You can see from Figure 6 that the investment in MAX5 is outweighed by the number of
reports that can be generated over that time frame.
Tests for web-based application throughput and response time
In this test, our goal was to determine how much MAX5 improves the performance of an
OLTP application that accesses a large database.
We used an OLTP application that used a WebSphere eXtreme Scale data grid as an inline
cache, meaning that the application directed all read/write requests to the grid. All equipment
was production-quality. If the data referenced in the request was not already in the grid, a
special database loader brought the data in by accessing the database. This action incurs
substantial processing overhead because the database frequently accesses a disk to obtain
the data.
Average Complex Cost Per
Report
(Lower is Better)
31% Cheaper
$19.39
$13.43
$0.00
$5.00
$10.00
$15.00
$20.00
$25.00
$
cost
per
report
40% Cheaper
$0.31
$0.19
$0.00
$0.05
$0.10
$0.15
$0.20
$0.25
$0.30
$0.35
$
cost
per
report
64% Cheaper
$0.52
$0.19
$0.00
$0.10
$0.20
$0.30
$0.40
$0.50
$0.60
$
cost
per
100
reports
Average Intermediate Cost Per
Report
(Lower is Better)
Average Simple Cost Per
Report
(Lower is Better)

Test setup and configuration
The server used in this test was a production-quality System x3690 X5 running Red Hat
Enterprise Linux 6 and WebSphere eXtreme Scale 7.1. The DB2 system was on a separate
IBM server with a storage system that supported 32 disks. Transactions were generated by a
remote client system. We set up two memory configurations:
򐂰 With MAX5: 256 GB RAM internal to the System x3690 X5, plus 256 GB RAM installed in
the MAX5 and attached to the System x3690 X5, for a total of 512 GB RAM
򐂰 Without MAX5: 256 GB RAM internal to the System x3690 X5 (called the base system)
This test configuration is shown in Figure 7. WXS represents WebSphere eXtreme Scale.
Figure 7 Test configuration that shows the System x3690 X5 with MAX5 and the 350 GB database
Test day
The first part of the experiment was to run WebSphere eXtreme Scale on the System x3690
X5 with MAX5 attached, thus providing a system with 512 GB RAM. We created a
WebSphere eXtreme Scale grid that uses 450 GB of RAM (45 containers with 10 GB each).
The database was approximately 350 GB in size and fit completely in the grid. During the
experiment, all data reads were satisfied from the grid.
The experiment contained two separate run scenarios, each with a different workload. In one
test scenario, the workload consisted of 80% read requests and 20% write requests. On write
requests, WebSphere eXtreme Scale first updated the grid and then synchronously wrote the
data to the database. In the second scenario, the workload consisted of 100% reads.
In the second part of the experiment, the tests were repeated on the System x3690 X5
without the MAX5. This configuration reduced the amount of memory by one-half, leaving 256
GB in the system. We reduced the grid on that system by half as well, thus creating a grid that
used 225 GB of RAM (45 containers with 5 GB each). There was not enough memory for the
entire 350 GB database. As a result, some requests to the grid failed, and WebSphere
eXtreme Scale read the data into the grid from the database. This state incurred the
processing overhead of a remote call and possibly a disk access. To ensure that the grid
System x3690
16 cores
256 GB
Application
Database
WXS
Container
WXS
Container
. . .
WXS
Container
WXS
Container
Client
Machine
Client
Machine
Database
Server
MAX5
Added 256 GB
. . .
WebSphere eXtreme Scale 7.1

always had space available for newly requested data, we used the WebSphere eXtreme
Scale built-in time-to-live evictor. The evictor removed any object in the grid not referenced for
one hour.
Results
Based on actual measurements, throughput with MAX5 was 2.4 - 4.9 times that of the system
without MAX5. Response times were 60% - 80% less than the response times of the base
system. Extrapolated results also showed that it can cost 49% more for a system without
MAX5 to match the performance of a system with MAX5.
MAX5 increases throughput and reduces response time
For the workload scenario of 80% reads and 20% writes, throughput on the MAX5 system
was 2.4 times greater than the throughput of the base system, as shown in Figure 8.
Response time on the MAX5 system was 60% less than on the base system, as shown in
Figure 9.
Figure 8 Throughput: 80% read 350 GB database (larger is better)
Figure 9 Response time: 80% read 350 GB database (smaller is better)
10280
4166.33
0
2000
4000
6000
8000
10000
12000
Throughput
Transactions/sec
With MAX5 Without MAX5
39.25
96.33
0
20
40
60
80
100
120
Response time
Milliseconds

For the workload scenario of 100% reads, throughput on the MAX5 system was 4.9 times
greater than the throughput of the base system, as shown in Figure 10. Response time on the
MAX5 system was 80% less than the response time on the base system, as shown in
Figure 11.
Figure 10 Throughput: 100% read 350 GB database (larger is better)
Figure 11 Response time: 100% read 350 GB database (smaller is better)
MAX5 reduces the total cost of acquisition
The STL team examined how much hardware had to be added to the base system to achieve
the same 2.4 times gain in throughput observed in the MAX5 system for the 80% read test
case. A comparable system without MAX5 is shown in Figure 12 on page 13.
The cost of this new system can be as much as 49% more than the cost of the MAX5 system
that was used in our tests. Table 1 on page 13 shows our purchase requirement calculations.
Values in the table are list prices that were published on the IBM website when the analysis
was conducted in mid-2011. They might not be representative of current list prices. You can
discuss your current environment with your IBM representative to determine your savings with
MAX5 at the current prices. A large portion of the increased cost is from the purchase of
additional WebSphere eXtreme Scale licensing. WXS represents WebSphere eXtreme Scale.
21957.85
4473.485
0
5000
10000
15000
20000
25000
Throughput
Transactions/sec
17.875
89.75
0
20
40
60
80
100
Response time
Milliseconds

Figure 12 A system without MAX5 (left) needs to provide the same performance as the system with MAX5 (right)
Table 1 Details of purchase requirement calculations
Hardware/software Cost Quantity
without MAX5
Total Quantity
with MAX5
Total
System x3690 X5
(includes 256 GB RAM)
$17,364.00 2 $34,728.00 1 $17,364.00
MAX5 $7,495.00 0 $0.00 1 $7,495.00
256 GB RAM $12,000.00 0 $0.00 1 $12,000.00
Disks $202,856.00 $202,856.00
Red Hat Enterprise Linux $0.00 2 $0.00 1 $0.00
WebSphere eXtreme Scale $236,800.00 2 $473,600.00 1 $236,800.00
Total $711,184.00a
a. Values in the table are list prices that were published on the IBM website when the analysis was conducted in
mid-2011. They might not be representative of current list prices. You can discuss your current environment with
your IBM representative to determine how much MAX5 can save you at the current prices.
$476,515.00a
System x3690 X5
16 cores
256 GB
WXS
Container
WXS
Container
MAX5
256 GB
WXS
Container
WXS
Container
WXS
Container
WXS
Container
. . .
WXS
Container
WXS
Container
. . .
System x3690 X5
16 cores
256 GB
System x3690 X5
16 cores
256 GB
Application
Database
Application
Database
Client
Machine
Client
Machine Client
Machine
Client
Machine
. . . . . .

Summary
Our tests demonstrated that MAX5 can offer significant benefits to your organization:
򐂰 Response time can be decreased for business analytics and data warehouse workloads
by the addition of MAX5. In our tests, the ability to cache all of the data in memory cut the
response time from 1.5 - 2.8 times.
򐂰 Throughput can be increased by using MAX5 for business intelligence workloads. In our
week-long test, we produced twice as many reports with MAX5 than without. Therefore,
over a three-year period, the cost of reports can be 31% - 64% lower with the addition of
MAX5.
򐂰 MAX5 can provide significant performance gains for transaction processing applications
with large databases. With MAX5, throughput was 2.4 - 4.9 times greater than that of the
test system without MAX5. Similarly, MAX5 enabled response times that were 60% - 80%
less than the response times on the system without MAX5.
򐂰 MAX5 technology can reduce your total cost of acquisition because it adds memory only,
rather than additional hardware, software, and the associated operating costs.
򐂰 MAX5 can increase performance at a lower cost of acquisition to meet your business
needs. Our tests showed that it can cost 49% more for a system without MAX5 to match
the performance of a system with MAX5.
All testing was completed by using production-quality equipment. Test results are based on
actual measurements. Estimations of future results are based on extrapolations from actual
measurements. MAX5 technology is one of the features of the eX5 product line that puts it in
a class by itself. It is an optional unit for the IBM eX5 line of servers.
For more information
For more information about the products referenced in this publication, visit these websites:
򐂰 The IBM Sales team:
http://ibm.com/systems/x/
򐂰 IBM MAX5:
– The MAX5 Advantage: How IBM System x MAX5 Benefits Microsoft SQL Server Data
Warehouse Workloads, REDP-4751:
– IBM Unveils Industry’s First Systems that Rewrite Economics of ‘Industry-Standard’
Computing, 2010:
http://www.ibm.com/press/us/en/pressrelease/29570.wss
򐂰 IBM System x3690 X5 server:
– IBM System x3690 X5: Ground-breaking two-socket server delivers outstanding
performance, memory and storage:
http://www.ibm.com/systems/x/hardware/enterprise/x3690x5/
– IBM eX5 Portfolio Overview: IBM System x3850 X5, x3950 X5, x3690 X5, and
BladeCenter HX5, REDP-4650:

򐂰 IBM System x servers:
– High density, high performance: IBM System x white paper:
http://ibm.com/systems/x/
򐂰 IBM eX5 Enterprise Systems:
– IBM eX5 enterprise systems: Breakthrough innovation for smarter computing:
http://www.ibm.com/systems/info/x86servers/ex5/
򐂰 IBM Systems home page:
– IBM servers, storage, and solutions for your business:
http://www.ibm.com/systems/
򐂰 WebSphere eXtreme Scale:
– Product information page:
http://www.ibm.com/software/webservers/appserv/extremescale/
򐂰 DB2:
– Product information page:
http://www.ibm.com/software/data/db2/linux-unix-windows/
The team who wrote this paper
This paper was produced by a team of specialists from around the world working at the
International Technical Support Organization, Poughkeepsie Center.
Tim Bohn is an IBM Senior Certified IT Specialist and Open Group
Master Certified IT Specialist with the IBM Software Group,
Strategy and Technology Lab (STL). His current focus is the IBM
System x platform hardware and software. Before joining STL in
2010, Tim was the Global Community Practice Leader for the IBM
Rational® Systems Development Community, where he helped
clients with systems and software development practices. Prior to
joining IBM Rational Software, Tim worked as a software and
systems engineer for 15 years. Tim holds a Bachelor’s and Master’s
degree in Computer Science from USC. He is a yearly guest
lecturer at USC and UCLA.

Dr. Howard Operowsky is a Senior Technical Staff Member in the
IBM STL where he evaluates server and middleware technology
and is currently focused on virtualization. Since joining IBM, he has
been manager or technical leader of a variety of IBM products,
including WebSphere Commerce architecture, collaborative
multimedia applications, IBM AIX/ESA® operating system
architecture, the design and development of compilers and
programming languages for high-end parallel systems, and
performance analysis of mainframe operating systems. He has
co-authored multiple patents in the area of multimedia applications
and is the recipient of an IBM Outstanding Technical Achievement
Award for work on operating system architecture. He received his
Ph.D. in Computer Science from the Courant Institute of
Mathematical Sciences at New York University. He is a Senior
Member of the Institute of Electrical and Electronics Engineers
(IEEE).
Karen Lawrence is a Technical Writer with the IBM ITSO team in
North Carolina. She has 25 years of experience in IT, with expertise
in application design, change management, and the software
development lifecycle (SDLC). She has worked with SMEs on
leading technologies in global, regulated enterprises in the areas of
storage systems, security, and disaster recovery, and the
applications to support these areas. Karen holds a Bachelor’s
degree in Business Administration from Centenary College in New
Jersey.
Thanks to the following people for their contributions to this project:
Many thanks to Randall W. (Randy) Lundin, Product Marketing - High End System x, for his
contributions and sponsorship of this project.
Thanks also to Linda Robinson, IBM Redbooks® graphics specialist, and to the editing team
from the International Technical Support Organization for their contributions to this project.
Now you can become a published author, too!
Here’s an opportunity to spotlight your skills, grow your career, and become a published
author—all at the same time! Join an ITSO residency project and help write a book in your
area of expertise, while honing your experience using leading-edge technologies. Your efforts
will help to increase product acceptance and customer satisfaction, as you expand your
network of technical contacts and relationships. Residencies run from two to six weeks in
length, and you can participate either in person or as a remote resident working from your
home base.
Find out more about the residency program, browse the residency index, and apply online at:
http://www.ibm.com/redbooks/residencies.html

Stay connected to IBM Redbooks
򐂰 Find us on Facebook:
http://www.facebook.com/IBMRedbooks
򐂰 Follow us on Twitter:
http://twitter.com/ibmredbooks
򐂰 Look for us on LinkedIn:
http://www.linkedin.com/groups?home=gid=2130806
򐂰 Explore new IBM Redbooks publications, residencies, and workshops with the IBM
Redbooks weekly newsletter:
https://www.redbooks.ibm.com/Redbooks.nsf/subscribe?OpenForm
򐂰 Stay current on recent Redbooks publications with RSS Feeds:
http://www.redbooks.ibm.com/rss.html

© Copyright International Business Machines Corporation 2012. All rights reserved.
Note to U.S. Government Users Restricted Rights -- Use, duplication or disclosure restricted by
GSA ADP Schedule Contract with IBM Corp. 19
Notices
This information was developed for products and services offered in the U.S.A.
IBM may not offer the products, services, or features discussed in this document in other countries. Consult
your local IBM representative for information on the products and services currently available in your area. Any
reference to an IBM product, program, or service is not intended to state or imply that only that IBM product,
program, or service may be used. Any functionally equivalent product, program, or service that does not
infringe any IBM intellectual property right may be used instead. However, it is the user's responsibility to
evaluate and verify the operation of any non-IBM product, program, or service.
IBM may have patents or pending patent applications covering subject matter described in this document. The
furnishing of this document does not give you any license to these patents. You can send license inquiries, in
writing, to:
IBM Director of Licensing, IBM Corporation, North Castle Drive, Armonk, NY 10504-1785 U.S.A.
The following paragraph does not apply to the United Kingdom or any other country where such
provisions are inconsistent with local law: INTERNATIONAL BUSINESS MACHINES CORPORATION
PROVIDES THIS PUBLICATION “AS IS” WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR
IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF NON-INFRINGEMENT,
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Some states do not allow disclaimer of
express or implied warranties in certain transactions, therefore, this statement may not apply to you.
Any performance data contained herein was determined in a controlled environment. Therefore, the results
obtained in other operating environments may vary significantly. Some measurements may have been made
on development-level systems and there is no guarantee that these measurements will be the same on
generally available systems. Furthermore, some measurements may have been estimated through
extrapolation. Actual results may vary. Users of this document should verify the applicable data for their
specific environment.
This information could include technical inaccuracies or typographical errors. Changes are periodically made
to the information herein; these changes will be incorporated in new editions of the publication. IBM may make
improvements and/or changes in the product(s) and/or the program(s) described in this publication at any time
without notice.
Any references in this information to non-IBM websites are provided for convenience only and do not in any
manner serve as an endorsement of those websites. The materials at those websites are not part of the
materials for this IBM product and use of those websites is at your own risk.
IBM may use or distribute any of the information you supply in any way it believes appropriate without incurring
any obligation to you.
Information concerning non-IBM products was obtained from the suppliers of those products, their published
announcements or other publicly available sources. IBM has not tested those products and cannot confirm the
accuracy of performance, compatibility or any other claims related to non-IBM products. Questions on the
capabilities of non-IBM products should be addressed to the suppliers of those products.
This information contains examples of data and reports used in daily business operations. To illustrate them
as completely as possible, the examples include the names of individuals, companies, brands, and products.
All of these names are fictitious and any similarity to the names and addresses used by an actual business
enterprise is entirely coincidental.
COPYRIGHT LICENSE:
This information contains sample application programs in source language, which illustrate programming
techniques on various operating platforms. You may copy, modify, and distribute these sample programs in
any form without payment to IBM, for the purposes of developing, using, marketing or distributing application
programs conforming to the application programming interface for the operating platform for which the sample
programs are written. These examples have not been thoroughly tested under all conditions. IBM, therefore,
cannot guarantee or imply reliability, serviceability, or function of these programs.

®
Redpaper™
This document REDP-4846-00 was created or updated on July 18, 2012.
Send us your comments in one of the following ways:
򐂰 Use the online Contact us review Redbooks form found at:
ibm.com/redbooks
򐂰 Send your comments in an email to:
redbooks@us.ibm.com
򐂰 Mail your comments to:
IBM Corporation, International Technical Support Organization
Dept. HYTD Mail Station P099
2455 South Road
Poughkeepsie, NY 12601-5400 U.S.A.
Trademarks
IBM, the IBM logo, and ibm.com are trademarks or registered trademarks of International Business Machines
Corporation in the United States, other countries, or both. These and other IBM trademarked terms are
marked on their first occurrence in this information with the appropriate symbol (® or ™), indicating US
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 http://www.ibm.com/legal/copytrade.shtml
The following terms are trademarks of the International Business Machines Corporation in the United States,
other countries, or both:
AIX/ESA®
Cognos®
DB2®
IBM®
Rational®
Redbooks®
Redpaper™
Redbooks (logo) ®
System x®
WebSphere®
X-Architecture®
z/OS®
The following terms are trademarks of other companies:
Intel Xeon, Intel, Intel logo, Intel Inside logo, and Intel Centrino logo are trademarks or registered trademarks
of Intel Corporation or its subsidiaries in the United States and other countries.
Linux is a trademark of Linus Torvalds in the United States, other countries, or both.
Microsoft, Windows, and the Windows logo are trademarks of Microsoft Corporation in the United States,
other countries, or both.
Java, and all Java-based trademarks and logos are trademarks or registered trademarks of Oracle and/or its
affiliates.
UNIX is a registered trademark of The Open Group in the United States and other countries.
Other company, product, or service names may be trademarks or service marks of others.
XSL03066-USEN-00

Add Memory, Improve Performance, and Lower Costs with IBM MAX5 Technology

More Related Content

What's hot (20)

Similar to Add Memory, Improve Performance, and Lower Costs with IBM MAX5 Technology (20)

More from IBM India Smarter Computing (20)

Add Memory, Improve Performance, and Lower Costs with IBM MAX5 Technology