Today I had the pleasure of joining Tom Krazit on stage at the Gigaom Structure’14 conference to share Intel’s vision of the data center in support of the growth of the digital services economy. We are in the midst of a bold industry transformation as IT evolves from supporting the business to being the business. This transformation and the move to cloud computing calls into question many of the fundamental principles of data center architecture. Two significant changes are the move to software defined infrastructure (SDI) and the move to scale-out, distributed applications. The speed of application development and deployment of new services is rapid. The infrastructure must keep pace. It must move from statically configured to dynamic, from manually operated to fully automated, and from fixed function to open standard.
As we have many times in our history, Intel is embracing this transformation and driving technology innovation to re-architect today’s data centers for the future.
As a first step, we start with a commitment to deliver the best technology for all data center workloads – spanning servers, network and storage. We started by augmenting our industry leading, general purpose Xeon processors with workload-optimized products, such as our Atom SoC processors for lightweight web-hosting, Xeon Phi for highly parallel processing, and the new Xeon D SoC series for hyperscale environments.
I have shared details of how we extended our products even beyond these workload-optimized solutions, delivering 15 custom products last year to meet specific needs of the end customers, including Ebay and Facebook. Today I disclosed that our development pipeline for custom solutions is growing, with more than twice the number of products planned for 2014.
But what we find even more exciting is our next innovation in processor design that can dramatically increase application performance through fully custom accelerators. We are integrating our industry leading Xeon processor with a coherent FPGA in a single package, socket compatible to our standard Xeon E5 processor offerings.
Why are we excited by this announcement? The FPGA provides our customers a programmable, high performance coherent acceleration capability to turbo-charge their critical algorithms. And with down-the-wire reprogramability, the algorithms can be changed as new workloads emerge and compute demands fluctuate. Based on industry benchmarks FPGA-based accelerators can deliver >10X performance gains. By integrating the FPGA with the Xeon processor, we estimate that customers will see an additional 2X in performance thanks to the low latency, coherent interface.
Our new Xeon+FPGA solution provides yet another customized option, one more tool for customers to use to improve their critical data center metric of “Performance/TCO”. It highlights our commitment to delivering the very best solutions across all data center workloads and our passion to lead in the transformation of the industry to cloud services.
Our innovation extends beyond silicon as we collaborate with industry partners to accelerate the move to SDI. Our early work with Facebook – as part of OCP – to define a rack-scale architecture (RSA) is gaining momentum with cloud service providers, telco service providers and hosters. The current data center optimization point is a single node and scale comes through deploying more nodes. The optimization point is moving to the rack – multi-node. The move to pooled, multi-node solutions – configurable and composable by software – will deliver the next step function in datacenter efficiency. And we will lead with the distributed compute, memory and switching technology required to realize the RSA reference architecture.
Rapid adopters of SDI are the network operators. With the continued rise in network capacity demand and the desire to rapidly deploy and monetize new services, the move to network function virtualization (NFV) is compelling. As a member of the ETSI NFV Forum, Intel contributed to the creation of the NFV spec and the development of the nine use cases. There are now 20 PoCs in flight with carriers from around the world, running on Intel architecture, up from nine just four months ago. Historically it has taken years for the service providers to deploy new services from initial concept. With the decoupling of the hardware and software through virtualization, the industry expectation is that service deployment time can reduce form years to months and eventually to minutes. Telefonica has announced that they will convert 30 percent of its network to NFV by 2016. With over 10 years of technology innovation in server virtualization and a legacy in building technology ecosystems, we are leading the way in the network transformation to SDI.
Finally, we are placing deep investment in unlocking the vast amounts of structured and unstructured data, working with industry leaders to speed delivery of modern analytics solutions. Our recent announcement of our collaboration and partnership with Cloudera is a clear example of our objective to enable the economic value enterprises can obtain through big data insight. We are committed to accelerating enterprise feature integration, ease the deployment and deliver optimized solutions for data analytics.
As we have seen many times during our history, transformation creates both opportunities and threats. Anyone in our industry trying to cling to the legacy world will be left behind. We see the move to cloud services and software defined infrastructure as a tremendous opportunity and we are seizing this opportunity. We are investing billions of dollars every year in data center R&D to ensure we meet the evolving needs of the customers. We invite others in the industry to join us in delivering to the vision of the data services economy.
source: Diane Bryant/Intel