Slashdot Mirror
NVIDIA Predicts 570x GPU Performance Boost
Gianna Borgnine writes "NVIDIA is predicting that GPU performance is going to increase a whopping 570-fold in the next six years. According to TG Daily, NVIDIA CEO Jen-Hsun Huang made the prediction at this year's Hot Chips symposium. Huang claimed that while the performance of GPU silicon is heading for a monumental increase in the next six years — making it 570 times faster than the products available today — CPU technology will find itself lagging behind, increasing to a mere 3 times current performance levels. 'Huang also discussed a number of "real-world" GPU applications, including energy exploration, interactive ray tracing and CGI simulations.'"
In other news, ATI is selling their 4870 series cards for $130 on newegg, which are twice as fast as an Nvidia 9800GTS which is the same price (at least on Left 4 Dead, Call of Duty, and any other game that matters). ATI is blowing Nvidia out of the water in terms of performance per dollar and will continue to do so through at least the middle of next year. See here:
http://www.tomshardware.com/charts/gaming-graphics-cards-charts-2009-high-quality/benchmarks,62.html
Yeah, I'd be making outrageous statements too if I were Nvidia.
moox. for a new generation.
The prediction is complete nonsense. It assumes that CPU processors only get 20% faster per year (compounded). That would only be true if they did not add more cores to the CPU. And finally GPUs are hitting the same thermal/power leakage wall that CPUs hit several years ago - they will at best get faster in lock step with CPUs.
A GPU is not a general purpose processor, as is a CPU. It is only good at performing a large number of repetitive single precision (32 bit) floating point calculations without branching. Double precision (64 bit) calculations - double in C speak - is 4 times slower than single precision on a GPU. And the second you have an "if" in GPU code, everything grinds to a halt. Conditions effectively break the GPU SIMD (single instruction multiple data) model and bring the pipeline to a halt.
The IEEE figures that semiconductor tech will be at the 11nm level around 2022. Intel and Nvidia both claim that they'll be significantly further along the path than the IEEE's roadmap. Maybe they're right, and I hope they are, but there are some very significant problems that appear as the process shrinks to that level.
You can never go home again... but I guess you can shop there.
Even generously assuming they'd achieve an 8x die shrink, that'd need to be producing chips with a 41000mm^2 die area. (Their current chips are already the biggest at 576mm^2.)
The GeForce 9 series was a rebrand/die shrink of GeForce 8, but the GTX 200 series has some major improvements under the hood:
* Vastly smarter memory controller including better batching of reads, and the ability to map host memory into the GPU memory space
* Double the number of registers
* Hardware double precision support (not as fast as single, but way faster than emulating it)
These sorts of things probably don't matter to people playing games, but they are huge wins for people doing GPU computing. The GTX 200 series has also seen a minor die shrink during the generation, so I don't know if the next generation will be more of a die shrink or actually include improved performance. (Hopefully the latter to keep up with Larrabee.)
Simple. GPUs are already massively parallell in a way that is actually usable. While the gigahertz-race has pretty much stopped, the transistor-race is still on, but getting the most out of a multi core CPU for a single application is non-trivial. GPUs however are a completely different ballgame, where the performance of the card pretty much scales with the number of shader cores.
You mean like the Tesla?
No, that won't do. The NVIDIA architecture (which is shared between Tesla and graphic cards) is 32-bit, meaning that it can only flat-address 4GB of RAM tops. The more sophisticated Tesla solutions are essentially built from clusters of Tesla cards, each with its own 4GB of RAM tops. Separate memory spaces means expensive memory transfers to share data between the cards, which is not an issue if you can get good domain decomposition, but is a BIG issue if you cannot.
The revolution for HPC on GPUs would be a 64-bit GPU architecture.
Proper support for doubles and possibly even long doubles would be a plus, for applications that need it.
"I'm never quite so stupid as when I'm being smart" (Linus van Pelt)
Well, it comes down to simple math. For the performance to get to 570-fold more than what it is now, in the same style package, either:
Both seem highly unlikely.
It's not a linear relationship.
"GPUs however are a completely different ballgame, where the performance of the card pretty much scales with the number of shader cores."
But only to the degree that your problem maps to that level of parallelization. There are many problems that do not perform well on the GPU.
But that's why there's a specific problem right in the name.
For the rest of us of course 570% increase is 5.7X faster.
It seems the rest of us don't understand what a "percent increase" means, either.
(hint: 570% increase == 6.7X)
sic transit gloria mundi
Where is my math wrong?
This isn't about math (well maybe a little) as much as it is about wording. Basically, the difference between "as fast/increasing/the speed" and "faster/increase". The first is a multiplicative action, while the second is an additive. So if you say, 100% as fast, you are basically saying a * 100% = a. While if you are saying 100% faster you are saying a + a*100% = 2*a. Now looking at the posts. You started by this assertion.
seeing 570% increase and going
OK. A 570% Increase. That would be a+a*570% = 6.7*a.
570% increase is 5.7X faster.
increase/faster. Good.
GPUs increasing 5.7X
Ouch. Here is the mistake. Using increasing. Suddenly you have downgraded the new anticipated speed from 6.7*a to a*5.7=5.7*a.
Now, in your second post you used the multiplicative action straight through. But that is pretty much the opposite of the first post that used the additive action except for a single time at the end.
Also, as a general advice, use X (times) only to represent a multiplicative action. That is what the general meaning of the word is, so it can be confusing when you here it describing an additive action. Not really wrong, but confusing.