Intel Xeon Processors E5 Achieve Fastest Adoption, Announcing Xeon Phi Co-Processors

[btarunr]

The Intel Xeon processor E5-2600 product family reached a new supercomputing milestone as the fastest adopted new processing technology to power 44 systems, including 3 Petascale-class supercomputers on the 39th edition of the Top500 list announced today.

The "SuperMUC" supercomputer at LRZ in Germany, which ranked fourth on the list, delivers 2.9 PetaFLOPs of performance, making it the most powerful in Europe, as well as the largest installation based on the new Intel Xeon processors E5 family.



At the International Supercomputing Conference, Intel Corporation also announced that Intel Xeon Phi is the new brand name for all future Intel Many Integrated Core Architecture (Intel MIC architecture) based products. Available by the end of 2012, the first generation of Intel Xeon Phi product family (coprocessors codenamed "Knights Corner") will complement the existing Intel Xeon processor E5-2600/4600 product families and deliver new levels of performance for highly parallel workloads. While the first generation primarily targets high performance computing (HPC), future generations of Intel Xeon Phi products will also address enterprise datacenters and workstations.

"The Intel Xeon processor E5 family is powering exponential performance gains in high performance computing and we're proud that it is having such a profound impact on the industry as demonstrated by its presence inside 44 of the Top500 supercomputers," said Raj Hazra, Intel Corporation VP and general manager of the Technical Computing at Data Center and Connected Systems Group. "As we add Intel Xeon Phi products to our portfolio, scientists, engineers and IT professionals will experience breakthrough levels of performance to effectively address challenges ranging from climate change to risk management. This is the next step of Intel's commitment to achieve exascale-level computation by 2018, and create a unique technology category that delivers unprecedented performance for today's highly parallel applications."

First Intel Xeon Phi Coprocessor Due This Year
Intel disclosed new technical details of the first commercially available product from its Intel Xeon Phi product family, a coprocessor codenamed "Knights Corner." In addition to delivering breakthrough performance for highly parallel applications, Intel Xeon Phi coprocessor's ease of use is bolstered by the benefits of familiar programming models, techniques and developer tools available with Intel architecture. With greater use of parallel CPU code, software companies and IT departments do not have to retrain developers on proprietary programming models associated with accelerators.

Beyond its compatibility with x86 programming models, the Intel Xeon Phi coprocessor will be visible to applications as an HPC-optimized, highly-parallel, separate compute node that runs its own Linux-based operating system independent of the host OS. This feature allows more flexibility when implementing cluster solutions that are not available with alternative graphics accelerator-based technologies.

Made with Intel's innovative 22nm, 3-D tri-gate transistors, the Intel Xeon Phi coprocessor, available in a PCIe form factor, contains more than 50 cores and a minimum of 8GB of GDDR5 memory. It also features 512b wide SIMD support that improves performance by enabling multiple data elements to be processed with a single instruction. Last year Intel showed a live demonstration of the single Knights Corner coprocessor delivering over 1 TeraFLOPs (1 trillion floating point operations per second) of double precision real life performance, as measured by DGEMM. At ISC'12 Intel demonstrated the same effective performance of more than 1 TeraFLOPs per node but measured by the industry standard benchmark Linpack (Rmax)1. By comparison, in 1997, it took more than 9000 Intel® Pentium® processors inside the ASCII RED* supercomputer to break the 1 TeraFLOPs barrier.

While initial production product shipments are planned for the second half of 2012, Intel has announced that the first Intel Xeon Phi coprocessor-based development cluster is up and running and ranked 149th on the Top500 list, delivering 118 TFLOPs of performance.

The Intel Xeon Phi coprocessor has strong industry support, with 44 manufacturers including Bull, Cray, Dell, HP, IBM, Inspur, SGI and NEC committed to including it in their system roadmaps.

"We are very excited to announce that our next-generation supercomputer code-named 'Cascade' will be available with the Intel Xeon Phi coprocessors, giving Cray customers the ability to push the limits of research and discovery," said Peg Williams, Cray's senior vice president of high performance computing systems. "Our Cascade system will feature some of the most advanced and highly innovative HPC technologies ever put into a Cray supercomputer. Combining these features with industry-leading Intel Xeon processors and the new Intel Xeon Phi coprocessors will result in a very compelling system for HPC centers around the world."

Due to power-on in early 2013, the first Petascale class supercomputer powered by a combination of the Intel Xeon processor E5 family and Intel Xeon Phi coprocessors will be "Stampede." Intel expects a growing number of Petascale-class machines in the next year that take advantage of the Intel Xeon Phi coprocessors' programmability and performance efficiency.

Race to Exascale fuels growth and investments in HPC market
Providing a two week weather forecast with the same level of accuracy as in a 48-hour forecast, or mapping the human genome within 12 hours at less than $1000 as compared to the current two weeks at $60,000, are two examples of the many challenges that HPC will address with more compute capacity. The insatiable need for performance has driven the tremendous growth of HPC processor shipments over the last five years, and Intel predicts this growth will reach more than 20 percent CAGR in the next 5 years. Intel forecasts2 that the most powerful supercomputer in 2013 will feature the amount of CPUs greater than 1 percent of its own server CPU shipments in 2011.

Reaching Exascale levels of performance requires innovation across the architecture and fabric within HPC. To deliver on its promise to reach Exascale performance by 2018, Intel made significant investments in areas that will accelerate its capabilities to deliver this performance in the future. The recent acquisition of Infiniband and interconnect assets from QLogic and Cray present key areas for Intel to innovate as the company eliminates the bottlenecks in delivering scalable Exascale-class platforms in the future.

TOP500 Supercomputers
More than 74 percent (372 systems) of the supercomputers on the 39th edition of the Top500 list are powered by Intel processors. Of those systems making their first appearance on the list, Intel-powered systems account for more than 77 percent. The complete report is available at www.top500.org.

[qwerty_lesh]

Is that you, Larabee?

[Completely Bonkers]

Don't be shy, come out of the closet!

[Mussels]

Originally Posted by qwerty_lesh

Is that you, Larabee?

exactly my thoughts

[Ikaruga]

We badly need a review of this asap (please).

[Jizzler]

But will it play Crysis FOLD?! Get one over the Pande Lab now!

[RejZoR]

I'm confused. The stand alone picture has no fan and the one where that guy is holding it has one...

[repman244]

Originally Posted by RejZoR

I'm confused. The stand alone picture has no fan and the one where that guy is holding it has one...

It could be 2 different versions, a fan is reasonable for normal chassis but pointless when you have many of them or having them in a 2U chassis.



I'm really interested of performance/power ratio of it.

[blanarahul]

Originally Posted by repman244

It could be 2 different versions, a fan is reasonable for normal chassis but pointless when you have many of them or having them in a 2U chassis.



I'm really interested of performance/power ratio of it.

Should be reasonable for the market Intel is targeting.

[AlienIsGOD]

It is technically Larrabee, but refined http://en.wikipedia.org/wiki/Intel_MIC

From Wikipedia's Larrabee entry "The project to produce a GPU retail product directly from the Larrabee research project was terminated in May 2010.[6] The Intel MIC multiprocessor architecture announced in 2010 inherited many design elements from the Larrabee project, but does not function as a graphics processing unit; the product is intended as a co-processor for high performance computing. The prototype card is named Knights Ferry, a production card built at a 22nm process named Knights Corner is planned for production in 2012 or later."

[Inceptor]

Originally Posted by Ikaruga

We badly need a review of this asap (please).

No we don't.
An HPC coprocessor is far outside of the bounds of normality and financial feasibility for nearly all the users of this forum.
Do you have enough money to waste on something that most likely costs much more than a typical Tesla card?

[Protagonist]

Easy to say i want, but wouldn't know what to do with it, all that power... No One Man Should Have All That Power.

[v12dock]

Cool logo but AMD could come out with a similar but faster product with GCN

[Static~Charge]

Originally Posted by v12dock

Cool logo but AMD could come out with a similar but faster product with GCN

The key word in that statement being "could". If/when it actually happens, it will be news.

[FordGT90Concept]

Originally Posted by qwerty_lesh

Is that you, Larabee?

Pretty sure it is but not capable of performing as a GPU.

Originally Posted by Inceptor

Do you have enough money to waste on something that most likely costs much more than a typical Tesla card?

The cost is likely to be similar. It is competing with Tesla for HPC market control.

I want to know how many x86 cores it has and at what clockspeed.

[blibba]

Originally Posted by Jizzler

But will it play Crysis FOLD?! Get one over the Pande Lab now!

The folding performance will be incredible.

[Jizzler]

Originally Posted by FordGT90Concept

Pretty sure it is but not capable of performing as a GPU.

It's close enough for me Carmack and Sweeny have been playing around with ray-tracing recently with good results, and Daniel Pohl has been researching ray-tracing in games since 2004. Only a matter of time til someone starts working on a complete engine.

Originally Posted by blibba

The folding performance will be incredible.

Yeah, hope to see a FAH for PHI implementation at some point! Even if they're at Telsa prices initially, there's already a market given how many people throw down $3500 for a 4-way Opteron Folding setup.

[ensabrenoir]

Originally Posted by v12dock

Cool logo but AMD could come out with a similar but faster product with GCN

.........is ....he serious........

[Ikaruga]

Originally Posted by Inceptor

No we don't.
An HPC coprocessor is far outside of the bounds of normality and financial feasibility for nearly all the users of this forum.
Do you have enough money to waste on something that most likely costs much more than a typical Tesla card?

I'm Enthusiast, and I had an impression techpowerup is somehow related to that. That's a brand new high-end hardware right there, and if you are not interested how that thing can perform, that doesn't mean that others can't have a different view.

[trickson]

Originally Posted by v12dock

Cool logo but AMD could come out with a similar but faster product with GCN

Now that is funny!

[AlienIsGOD]

Originally Posted by Ikaruga

I'm Enthusiast, and I had an impression techpowerup is somehow related to that. That's a brand new high-end hardware right there, and if you are not interested how that thing can perform, that doesn't mean that others can't have a different view.

True, but if you want a review of this product, just go find a HPC related tech site, I'm sure that reviews or benchmarks can be found when the product is released.

[T4C Fantasy]

someone call up mydog, tell him to buy one and tell him to fold and watch his ppd, thats a benchmark ill wait for.

[Ikaruga]

Originally Posted by AlienIsGOD

True, but if you want a review of this product, just go find a HPC related tech site, I'm sure that reviews or benchmarks can be found when the product is released.

I obviously did not expect TPU to review it, I was just really curious about the numbers and the possibilities

[Drone]

Wow install a million of those things and you have exaflop.

[Morgoth]

i want this to be a part of my evga w555 setup