Paper

Scale Up Performance with
Intel and IBM
Not all workloads scale out well. Optimizing
heavy database workloads or virtualizing
tier-1 enterprise resource planning (ERP)
or customer relationship management
(CRM) workloads requires scaling up the
performance of your servers. Intel and IBM
provide the tools to do so on the hardware
and software levels.
Next-Generation Intel Xeon Processor
E7 Family: Beyond Incremental
The Intel® Xeon® processor E7 v2 family
delivers an average of two times improved
performance over the earlier generation
Intel Xeon processor E7 family.
3,4
But
intensive workloads also benet from
more memory, CPU scale, and reductions
to input/output (I/O) bottlenecks.
Scale Memory, Processors,
and Input/Output
The next-generation Intel Xeon processor
E7 v2 family supports up to three times
more memory than previous generation
Intel Xeon processors. The four-socket
Intel Xeon processor E7 v2 family can
support up to 6 TB of memory, whereas
the eight-socket versions can use up to 12
TB. This enables you to put large primary
databases directly in memory for orders of
magnitude increased performance.
You can also dynamically scale CPUs in the
next-generation Intel Xeon processor E7
v2 family, which natively uses two, four, or
eight sockets without any additional work.
And with third-party node controllers, you
can scale this up to 256 sockets.
Every sensor reading, every Web click, every social media post contributes to the
torrent of data rushing at you. Just getting a handle on the volume and variety of data
coming at you with ever-increasing velocity is challenging in itself, and that’s before
trying to analyze it. But the value of big data isn’t the data itself—it’s uncovering the
insights locked in your data that is important.
Even if you have your big data strategy in place, there’s still a lot of data to handle—and
more of it all the time. According to the Harvard Review of Business, over 2.5 exabytes
of data got created every day in 2012.
1
Put another way, in 2012, as much data was
created daily as existed in the entire world in 1986.
2
The physical limits of computing, such as the speed of accessing data from relational
databases, have created bottlenecks in data analysis. Complex calculations could take
hours or even days, and data could provide insights, but not always fast enough to
maximize value. All of this is changing with the next-generation of processors and
database software from Intel and IBM, which can help you extract near real-time
insights from critical data.
Unlock the Value of Your Data
with Intel and IBM Innovations
Extract near real-time insights from critical data using IBM* DB2* with
BLU Acceleration running on the Intel® Xeon® processor E7 v2 family
White Paper
Intel® Xeon® processor E7 v2
Big Data