What Intel Xeon E5 Processor Means to Intel IT

Add Your Comments

Ajay Chandramouly has more than 12 years of experience in the technology industry, and more than 9 years of experience at Intel Corporation. Ajay held a variety of IT, software and hardware engineering positions while at Intel, and also at the Lawrence Livermore National Laboratory. You can find him online at Ajayc47 on Twitter.

Ajay-ChandramoulyAJAY CHANDRAMOULY
Intel

At Intel IT, we see Intel products through the eyes of the end-user IT community. As an Intel customer and an early adopter of new Intel technologies, we are writing this blog to share our experiences and best practices with our peers in other enterprise IT organizations.

That’s the case with the new Intel Xeon Processor E5 Family. We have extensively tested and used the Xeon E5 processors in production environments, and are embarking on a major technology refresh based on the Xeon E5 architecture. In this post, I want to bring you our insights into the technology refresh and the new architecture.

Refresh Is All About Business Value

Technology refresh is central to delivering business value back to the corporation. It’s one of the keys to driving improved IT services at a lower cost.

At Intel IT, technology refresh has supported hundreds of millions of dollars in efficiencies. In one example of these efficiencies, we have achieved server consolidation ratios as high as 20:1 by refreshing with the latest-generation Intel Xeon servers—and gained all the corresponding savings on power, cooling, and licensing.

We’ve also achieved major gains on the cloud front. The private cloud infrastructure we have implemented here saved $9 million net to date and will save an expected $6 million annually over the next four years. This is strictly cash savings; the total doesn’t include savings from efficiencies gained through increased agility, which by the way was the biggest reason for implementing our private cloud. And we have identified that moving to a blades form factor helped improve our cloud total costs of ownership by 29 percent.

Technology refresh is also driving facilities savings. Investing in the latest and greatest Intel architecture has helped us avoid building new data centers to accommodate growth. Instead, we’re consolidating data centers. In 2010, we had 91 data centers. Today we have 87, and we have a data center strategy in place that should help us reduce that count by 35 percent over the next few years.  By investing in the latest IA, we’ve been able to reduce spending on facilities.

Focus on Performance Improvements

We tested servers based on the E5 architecture, and we’ve seen how it can increase server performance. In tests conducted in our microprocessor design environment, we saw up to a 55 percent performance improvement over the earlier Intel Xeon 5600 processor. These were tests running actual production silicon design workloads, as opposed to more theoretical benchmarks. Gains of this magnitude in our design shop can have a huge impact on time to market for Intel products.

As a result of these tests, we intend to refresh based on the new E5 servers and deploy them across our entire environment—including our design shop, our manufacturing operations, and our office and enterprise and services group. We expect the E5 family to become the predominant platform throughout our environment.

It’s worth pausing here to explain why we are deploying E5 first in our design environment. We don’t have unlimited resources and so we test and deploy the latest and greatest technology in design first because design is at the heart of what Intel does. Design requires ever-greater amounts of compute power and throughput to handles increasing complex microprocessors.

We also expect the E5 architecture to deliver big gains in our enterprise private cloud, which now runs production applications for most Intel business groups. The performance gains from servers based on the E5 architecture will help Intel IT meet its future goals of having a federated cloud that operates at 80 percent utilization across 13 data centers.

Refresh Doesn’t Stop at Servers

It’s important to note that the refresh process doesn’t begin and end with servers. It’s also critical to refresh across storage, networks, and facilities, to avoid bottlenecks arising beyond the server.

This broad approach to technology refresh is one of our keys to achieving our goal of 80 percent effective utilization of resources across our global data centers. If you refresh only servers, you could end up with storage or networking bottlenecks that prevent you from achieving your targeted utilization rates.

But this is about more than bottlenecks. At Intel IT, the storage refresh and optimization process has helped drive $9.2 million in savings, while upgrading to 10 Gigabit Ethernet helped reduce our network costs by up to 25 percent.

All of this, of course, provides only a glimpse of Intel IT and our experiences with the Xeon E5 architecture. To learn more, visit us as www.intel.com/it.

Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library.

Add Your Comments

  • (will not be published)