Tuesday, June 21st 2011

Intel Equipped to Lead Industry to Era of Exascale Computing

At the International Supercomputing Conference (ISC), Kirk Skaugen, Intel Corporation vice president and general manager of the Data Center Group, outlined the company’s vision to achieve ExaFLOP/s performance by the end of this decade. An ExaFLOP/s is quintillion computer operations per second, hundreds times more than today’s fastest supercomputers.

Reaching exascale levels of performance in the future will not only require the combined efforts of industry and governments, but also approaches being pioneered by the Intel Many Integrated Core (Intel MIC) Architecture, according to Skaugen. Managing the explosive growth in the amount of data shared across the Internet, finding solutions to climate change, managing the growing costs of accessing resources such as oil and gas, and a multitude of other challenges require increased amounts of computing resources that only increasingly high-performing supercomputers can address.
“While Intel Xeon processors are the clear architecture of choice for the current TOP500 list of supercomputers, Intel is further expanding its focus on high-performance computing by enabling the industry for the next frontier with our Many Integrated Core architecture for petascale and future exascale workloads,” said Skaugen. “Intel is uniquely equipped with unparalleled manufacturing technologies, new architecture innovations and a familiar software programming environment that will bring us closer to this exciting exascale goal.”

Paving the Way to Exaflop Performance
Intel’s relentless pursuit of Moore’s Law -- doubling the transistor density on microprocessors roughly every 2 years to increase functionality and performance while decreasing costs -- combined with an innovative, highly efficient software programming model and extreme system scalability were noted by Skaugen as key ingredients for crossing the threshold of petascale computing into a new era of exascale computing. With this increase in performance, though, comes a significant increase in power consumption.

As an example, for today’s fastest supercomputer in China, the Tianhe-1A, to achieve exascale performance, it would require more than 1.6 GW of power – an amount large enough to supply electricity to 2 million homes – thus presenting an energy efficiency challenge.

To address this challenge, Intel and European researchers have established three European labs with three main goals: to create a sustained partner presence in Europe; take advantage of the growing relevance of European high-performance computing (HPC) research; and exponentially grow capabilities in computational science, engineering and strategic computing. One of the technical goals of these labs is to create simulation applications that begin to address the energy efficiency challenges of moving to exascale performance.

Skaugen said there is the potential for tremendous growth of the HPC market. While supercomputers from the 1980s delivered GigaFLOP/s (billions of floating point operations per second) performance, today’s most powerful machines have increased this value by several million times. This, in turn, has increased the demand for processors used in supercomputing. By 2013 Intel expects the top 100 supercomputers in the world to use one million processors. By 2015 this number is expected to double, and is forecasted to reach 8 million units by the end of the decade. The performance of the TOP500 #1 system is estimated to reach 100 PetaFLOP/s in 2015 and break the barrier of 1 ExaFLOP/s in 2018. By the end of the decade the fastest system on Earth is forecasted to be able to provide performance of more than 4 ExaFLOP/s.

Intel MIC Architecture Software Development Momentum
The Intel MIC architecture is a key addition to the company’s existing products, including Intel Xeon processors, and expected to help lead the industry into the era of exascale computing. The first Intel MIC product, codenamed "Knights Corner,” is planned for production on Intel’s 22-nanometer technology that featuring innovative 3-D Tri-Gate transistors. Intel is currently shipping Intel MIC software development platforms, codenamed “Knights Ferry,” to select development partners.

At ISC, Intel and some of its partners including Forschungszentrum Juelich, Leibniz Supercomputing Centre (LRZ), CERN and Korea Institute of Science and Technology Information (KISTI) showed early results of their work with the “Knights Ferry” platform. The demonstrations showed how Intel MIC architecture delivers both performance and software programmability advantages.

“The programming model advantage of Intel MIC architecture enabled us to quickly scale our applications running on Intel Xeon processors to the Knights Ferry Software Development Platform,” said Prof. Arndt Bode of the Leibniz Supercomputing Centre. “This workload was originally developed and optimized for Intel Xeon processors but due to the familiarity of the programming model we could optimize the code for the Intel MIC architecture within hours and also achieved over 650 GFLOPS of performance.”

Intel also showed server and workstation platforms from SGI, Dell, HP, IBM, Colfax and Supermicro, all of which are working with Intel to plan products based on “Knights Corner.” “SGI recognizes the significance of inter-processor communications, power, density and usability when architecting for exascale,” said SGI CTO Dr. Eng Lim Goh. “The Intel MIC products will satisfy all four of these priorities, especially with their anticipated increase in compute density coupled with familiar X86 programming environment.”

TOP500 Supercomputers
The 37th edition of the Top500 list, which was announced at ISC, shows that Intel continues to be a force in high-performance computing, with 387 systems or more than 77 percent, powered by Intel processors. Out of all new entries to the list in 2011, Intel powered systems accounted for close to 88 percent. More than half of these new additions are based on latest 32nm Intel Xeon 5600 series processors which now alone power more than 35% of all systems in TOP500 list, three times the amount comparing to last year.

The semi-annual TOP500 list of supercomputers is the work of Hans Meuer of the University of Mannheim, Erich Strohmaier and Horst Simon of the U.S. Department of Energy’s National Energy Research Scientific Com
Add your own comment

49 Comments on Intel Equipped to Lead Industry to Era of Exascale Computing

#1
cadaveca
My name is Dave
TheMailMan78 said:
But will it run Crysis?
:slap:
Posted on Reply
#2
Thatguy
mamisano said:
I wonder if this will be able to compete with AMD's "Graphics Core Next" GPU which is supposed to be heavily developed with compute in mind.
Its going to try, and fail miserably at it.
Posted on Reply
#3
Completely Bonkers
x86 PCI "blade" aka larrabee-phoenix, is a great way to scale a PC/workstation/server for asymmetric computational power.

One CPU "horse" and 100 "mice". Perfect for 99% of real-world applications where CPU is not 90%+ all the time, but where you want lots of background threads, services, applications or VMs "ready".

But I have said it before, and will say it again, windows task scheduler is not yet up to scratch to handle this. MS needs to build a more sophisticated scheduler that understands asymmetric CPU scaling.
Posted on Reply
#4
St.Alia-Of-The-Knife
FordGT90Concept said:
CUDA vs x86. x86 = much better and virtually anyone can program for it without having to learn much new. x86 is also far more powerful in that it can handle logic processing much better than CUDA.
o i get it, just like those pentium 4s, ordinary x86 architecture like a CPU
Posted on Reply
#5
spine7
2 million homes? 1.6GW/2000000=0.8Watt per home
Posted on Reply
#6
mastrdrver
Until Intel can make drivers: I'll believe it when I see it.

If nVidia is around long enough, then they will be first. If they are not, then AMD will be. In either case Intel will be last. If I could put money on it, I would.
Posted on Reply
#7
ASharp
spine7 said:
2 million homes? 1.6GW/2000000=0.8Watt per home
1.6GW is 1.6 billion watts...not 1.6 million. It would equal 800W per home. :slap:
Posted on Reply
#8
a_ump
ASharp said:
1.6GW is 1.6 billion watts...not 1.6 million. It would equal 800W per home. :slap:
that really enough for a household? a normal one that is. I know a lot of us on here use half to almost that much through our comps, but 800w?.... That sounds like xbox, tv, a light per room, no ac, and when you want to use the microwave you better unplug yo frig lol
Posted on Reply
#9
slyfox2151
a_ump said:
that really enough for a household? a normal one that is. I know a lot of us on here use half to almost that much through our comps, but 800w?.... That sounds like xbox, tv, a light per room, no ac, and when you want to use the microwave you better unplug yo frig lol
average over the course of 1 days might be 800 watts per hour.....

lots of people only have a fridge turned on during work/sleep.
Posted on Reply
#11
Thatguy
Hayder_Master said:
Ohhh P2 is back! slot cpu's again
Don't forget p1 and early p3 stuff as well as some AMD styled slot processors !
Posted on Reply
#13
meran
i cant see 32core beat gpu with 512 cores or 3200 sps
Posted on Reply
#14
MicroUnC
TheMailMan78 said:
But will it run Crysis?
:roll::roll::roll:
Posted on Reply
#15
[H]@RD5TUFF
afw said:
Is that a radeon card with an intel sticker :laugh:
No.

This is what has had ATI and Nvidia, scared for some time now, as Intel is the leading develpoer in advanced CPU architecture, it's not a stretch to think they could and are looking into GPU market, the first steps of which we see CPU's with integrated GOU's ion the chip, this is just the next step.
Posted on Reply
#16
DrPepper
The Doctor is in the house
meran said:
i cant see 32core beat gpu with 512 cores or 3200 sps
There is more to it than just cores, it's all about the architecture.
Posted on Reply
#17
[H]@RD5TUFF
DrPepper said:
There is more to it than just cores, it's all about the architecture.
One of the reasons why Intel quad cores beat AMD hexa cores.
Posted on Reply
#18
TheMailMan78
Big Member
[H]@RD5TUFF said:
One of the reasons why Intel quad cores beat AMD hexa cores.
This is true in most cases.
Posted on Reply
#19
mastrdrver
[H]@RD5TUFF said:
No.

This is what has had ATI and Nvidia, scared for some time now, as Intel is the leading develpoer in advanced CPU architecture, it's not a stretch to think they could and are looking into GPU market, the first steps of which we see CPU's with integrated GOU's ion the chip, this is just the next step.
In which case I still stand by my statement. Until Intel can prove they can write drivers, ATI and nVidia are far, far from scared of this. It's more along the lines of this: :roll:
Posted on Reply
#20
FordGT90Concept
"I go fast!1!11!1!"
I doubt these things need much of any drivers (just a little something something to say hey, you got extra x86 processors over here). Also, I suspect these cards are going to sell for $2000+ a pop so it isn't like any of us are in the market for them. IBM, Cray, and other super-computer manufactureres will likely buy them up by the hundreds of thousands though.
Posted on Reply
#21
DaedalusHelios
mastrdrver said:
In which case I still stand by my statement. Until Intel can prove they can write drivers, ATI and nVidia are far, far from scared of this. It's more along the lines of this: :roll:
Basically Intel could rule the scientific research and discrete graphics card markets if they were willing to spend that much R&D money on it. They have deep pockets. But there isn't that much money to be made in the enthusiast and scientific markets versus family internet machines, and business solutions.
Posted on Reply
#22
MikeX
would you say this is the x86 version equivalent of CUDA?
Posted on Reply
#23
FordGT90Concept
"I go fast!1!11!1!"
It's the Intel version of NVIDIA Tesla. XD
Posted on Reply
#24
mastrdrver
FordGT90Concept said:
I doubt these things need much of any drivers (just a little something something to say hey, you got extra x86 processors over here). Also, I suspect these cards are going to sell for $2000+ a pop so it isn't like any of us are in the market for them. IBM, Cray, and other super-computer manufactureres will likely buy them up by the hundreds of thousands though.
DaedalusHelios said:
Basically Intel could rule the scientific research and discrete graphics card markets if they were willing to spend that much R&D money on it. They have deep pockets. But there isn't that much money to be made in the enthusiast and scientific markets versus family internet machines, and business solutions.
If they didn't need drivers (or very basic ones) then why do the nVidia and AMD Tesla and Firepro cards (respectively) cost so much?

Its because of drivers.

Intel wants the cast cow. Problem is they don't have the feed to feed them with.
Posted on Reply
Add your own comment