Nvidia: Next big supercomputer player?
Summary: Nvidia unveiled its next-generation graphics processing unit (GPU) architecture, dubbed Fermi, and announced a key supercomputing win with the Oak Ridge National Laboratory.The takeaway: Graphics and visualization are becoming key to scientific discoveries.
Nvidia unveiled its next-generation graphics processing unit (GPU) architecture, dubbed Fermi, and announced a key supercomputing win with the Oak Ridge National Laboratory.
The takeaway: Graphics and visualization are becoming key to scientific discoveries. And Nvidia could be a major player. Oak Ridge's supercomputer will be used for research in energy and climate change and is expected to be 10 times more powerful than today's fastest supercomputer (statement).
Fermi, announced at Nvidia's GPU Technology developer's conference (statement), is a new architecture designed to deliver computational GPUs. The game for Nvidia CEO Jen-Hsun Huang is to make his company more than just a graphics chip player.
Fermi will have tools for large data center deployments, 512 CUDA Cores and a host of other goodies including NVIDIA Parallel DataCache, which speeds up algorithims for physics solvers and raytracing.
CNET News' Brooke Crothers notes:
The Fermi chip integrates three billion transistors, about three times the number of transistors in Nvidia's most powerful graphics chip now on the market. In the future, the chip will also find its way into Nvidia's GeForce product line for PCs...The architecture would use both graphics processing units (GPUs) from Nvidia and central processing units (CPUs), according to Nvidia. Intel and Advanced Micro Devices, among others, make the CPUs.
The problem? Nvidia is a big battleground among analysts. If faces fierce competition from AMD. In addition it's unclear whether these big supercomputing announcements boost Nvidia's commercial standing.
The battleground was evident on Thursday among Wall Street analysts.
JMP Securities Alex Gauna says there was a lot of buzz around the Nvidia tech conference, which was filled to capacity. Gauna gushes:
We would characterize the energy of the event as booming, with attendance some 50% oversubscribed beyond event organizer expectations and participants spilling into multiple overflow rooms to witness the dramatic inaugural keynote that jumped off presentation screens in stunning 3-D. Testimonials emerging from the conference came from the echelons of scientific supercomputing, academic research, medical devices, and industrial design in addition to the standard cadre of multimedia and entertainment advocates. We also overheard the real exchanging of ideas and business prospect on the exhibition floor at the end of the day, versus the usual one-sided market pitches we have come to expect from technology forums. This groundswell of support and enthusiasm for a technology roadmap that did not even exist two years ago is stunning to us and we believe promises dramatic scientific, commercial, and virtualization breakthroughs in coming years. It also represents an opportunity for Nvidia to entrench itself as a major mainstream computational force in the electronics industry in the coming decade, in our view.
And then there's the other side of the Nvidia equation. J.P. Morgan analyst Shawn Webster downgraded Nvidia. He said that Nvidia is on track for its fourth quarter, however, "we are concerned on risk to Consensus estimates for calendar 2010/2011 and continued competitive pressure from AMD should drive under-performance." Webster writes:
AMD appears ahead of NVDA on product cycle, despite “Fermi” announcement yesterday. AMD recently launched its new family of graphics processors on 40 nanometer technology. Third party reviews have been positive on the family in terms of its performance, cost and power consumption. We analyzed the early reviews and estimate AMD's "5800" family of products improves performance roughly 50% versus its prior generation at lower power, but at a chip size which is only 18% larger (albeit on a new process technology). While Nvidia announced its new “Fermi” (aka “GT300”) GPU architecture at a conference yesterday, many details were missing and it is not available yet and we believe may not arrive until December or later, which adds risk the company will miss the holiday season and the launch of Windows 7.
Webster adds that the Fermi architecture could result in a "poor cost structure" relative to AMD.
In the end, the Gauna vs. Webster takes couldn't be more different. As usual, the truth is probably in the middle somewhere. Bottom line: Nvidia can be a supercomputing player, but it may not matter to the company's commercial standing.
Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.
Talkback
NVidia has a chance
CUDA was recently touted in a number of studio specific publications for computation of convolution reverbs, audio dynamics, EQ and such, all of which are in a far smaller dollar universe than climate change supercomputing requirements, but the fact that one can have the opportunity to such computing power even in a recording studio certainly has to be of consequence to NVidia.
I, for one, look forward to these developments and their more widely based applications than just the National Computer Lab type of implementation.
RWNorman
RE: Nvidia: Next big supercomputer player?
Where's the Consumer Apps For It?
Amd is already selling there 5850 and 5870 series with more on the way at a consumer price point and with current applications that will benefit from them.
I think that is why the one Analyst downgraded them since it is not something that is going to sell in the same volume or space as a high-end Video Adapter.
It would have to be one real killer app for most consumers to spring $8K (on top of the puchase of a high-end computer) to take advantage of the computing power.
See My Response "Broaden Your Horizons"
How about the best flight simulator ever?
Imagine seeing these setups going into even small colleges and universities and private research organizations. It's my belief that we are about to see a tremendous upswing in breakthroughs in scientific and technical research all over the planet.
Nvidia has a vision. I can see it, and I hope they are successful.
AMD/ATI is pushing that way as well. I also wish them success.
We're living in exciting times.
Consumer apps won't be the determining factor for the adoption of this tech
architecture has emerged in recent years, and
vendors have continued to keep up with these
advancements, that isn't about to change with
the introduction of this or any other viable
architecture that emerges in the future.
I also think these 'consumer apps' so to speak
are irrelevant. Most operating systems have
some kind of HAL that works in kernel mode,
meaning 'apps' aren't written to talk directly
to the cpu/hardware anyway.
RE: Nvidia: Next big supercomputer player?
Nothing has any value till enough people want it. How many will actually buy an $8000 video card, and how much will the computer cost that you will put it in.
On the green side. I am glad to see that they are building a supercomputer to investigate the climate change.. We really do need to find out if what the political gurus, and the save the earth fanatics have any credibility. What if they find out, it is caused by Volcanoes and cattle? Oh My God!
Broaden Your Horizons
Computers with supercomputing capabilities for under $10,000 a pop - compared to $250,000 for just basic supercomputer hardware - would have a very large market appeal. Every single scientist and engineer who works with fluid flow dynamics and n-body problems all over the whole world would stand in line for one - atmospheric scientists, astronomers, geologists, cosmologists, theoretical physicists, planetary scientists, aerospace engineers, nuclear engineers, satellite mission planners, combinatorial chemists, geneticists, universities, NASA ... the list goes on and on.
I wonder how Fujitsu and Cray are taking the news.
The problem is, it's not for the masses.
I don't game, I can get by nicely with my Phenom II x4 and ATI graphics card. I run at 3.6 on air and can do most anything that I need to do. How many people will be willing to throw $10K at a computer. I built mine for $650.
You're correct, it wasn't built for the masses.
The applications will come...
However, the prices will have to come down to a level where most people can afford them.
A super-computer such as described above can eventually be programmed to perform many functions to assist a human in everyday life. A system such as that can do your cooking, do your washing, take out your trash, replace your teacher, be your library, save all of your data and family medical information, analyze your perspiration and your blood and your pulse and your urine and saliva and lungs and heart rate and let you know of any possible medical problems, remember your TV and music preferences and then search and make it available when you desire, inform you about the weather, and inform you on the latest news and other important information, drive your car, keep track of your kids, and much more. All of that and more with just one "super-computer" and many applications.
That drive your car thing would be possible with an automatic guidance system but the information could be centralized at your home super-computer which would be in direct communication with the guidance system in your car.
Yeah... just dreaming for now.
All The House Stuff Might Be Convenient
;-)
Broaden Your Horizons
Computer companies are going to have to come to grips with the new consumer outlook - 'Less is More'. Most home/business computer's processing power exceeds what most consumers need.
This is great development for high end proffesionals and super high end gamers, but a 'no impact' announcement for the rest of us.
Supercomputer or not, GIGO still applies
The problem with the global warming science is not the absence of a supercomputer that can take in the data and analyze it and spit out results.
The problem is with the intentional fudging with the scientific principles, in every step of the way. When the data is tampered with or "selected" in order to arrive at a predefined desired result, it's not the fault of the computer. When the scientific model is designed to specifically work towards a desired result, then it's also not a problem with the computer. When the analysis yields the desired result from the chosen data and from a model designed to arrive at a predefined result, it's also not a problem with the computer.
If a super-computer were to be designed that would be a trillion times more powerful than any in existence right now, it still would yield the desired results if the data and the models were tampered with in order to deceive. .
The climate change science will not be "correct" until a system is developed which is able to determine that the "scientists" weren't coming into the research with an agenda. It is the agenda which renders as junk any global warming science.
Sorry to go off on a tangent, but you did (sorta) bring it up.
Now, about that "super-computer" from Nvidia. It and any others that are developed will be put to good uses. Some will triumph while others will fail. But, there is a need for something like that right now. Once the prices come down from the many thousands, even people at home will be able to afford them. Remember the high prices of the first iteration of the latest and most up-to-date Intels, from the 8086 to the 286, 386, 486, Pentiums and now the multi-core x86s? Each time, the initial price was prohibitive for most people except small business and larger ones.
I myself could think of a system which could use the power of such an Nvidia or AMD (or Intel) system. In fact, I might already be working on that system.
RE: Nvidia: Next big supercomputer player?
RE: Nvidia: Next big supercomputer player?
persoanlly i like the idea, and i too remeber when the average processor chip cost 1k and one in a pc box was closer to 3k, i remeber when that price droped to 2k and i bought my first one (that was around 20 years ago) and i remember building a pc for the first time 15 years ago for under ?200, the price of thease systems will drop to the same price margin in the next 10 years end start being commonplace in the average home (which are clearly the destined objectives by the design of the press release about them) and no doubt we will all end up with something of that ilk sitting at the end of the desk powering all the 3/4d models we can throw at it for the next generation of displays (you just know there gonna make those 3d projectors from the star wars films for home users). all in all very definatly a bonus for the supercomputer market and finally some real competition which will add agressive priceing options to force the likes of cray to get real in the cost of making/supporting one of them(lets face it the average pc nerd could build one if they could get the components) also this should mean more discoveries like sequencing the human genome are possible in a fraction of the tiem it cost previously.
Um... no.
RE: Nvidia: Next big supercomputer player?
Keep it rolling.
I have read all the posts and the ones that point to failures kind of miss the point, if you do not try you will definitely fail (Narg).
We live in amazing and sometimes confusing computing times (even though I have had close to 25 years of computer experience) where all these ideas come from or are going to i.e. Cuda, cloud, virtualization etc etc.
They are all wonderful ideas, some are even astounding, but as an individual we have to be happy that these guy/girls are making the IT world a very interesting place to be.
Good work from both AMD and nVidia may put more pressure on the outragously vile Intel and M$ to toe the line and get back to fair marketing and R&D.
Keep up the great work.
P.S. Put me down for the Star Wars holo projector, or that Star Trek Holodeck before I am too old, and smell of urine.
in reltion to AMD
were putting a lot of time and resources into, it'll be
interesting to see how this development effects their own
roadmap.