Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

TOP500 at ISC’26: We have a New Number 1 Supercomputer

chipsandcheese.com

Post reply on HN
Print view View original

71–80 of 90 posts

Re: TOP500 at ISC’26: We have a New Number 1 Supercomputer

#71

Post by fragmede »

NitpickLawyer wrote: post #42
ziofill wrote: post #17

Earlier quoted context omitted.

Also, would those 550k Blackwell have good FP64 performance? How would one even compare them?

Yeah, that's true. GPU flops are really impressive in fp16, and more recently fp8 and fp4. When the 40xx GPUs came out, Tim Dettmers had a really cool blog looking into the numbers, and for fp8 one 4090 GPU had enough flops to match the best supercomputer somewhere in the 2000s, with an 8x 4090 build being top for a few years more. It's insane. But it gets nowhere close on fp64, where most of the physics simulations…

Other than weather simulations and nuclear explosions, what other supercomputer tasks are out there?

Re: TOP500 at ISC’26: We have a New Number 1 Supercomputer

#72

Post by NitpickLawyer »

fragmede wrote: post #71
NitpickLawyer wrote: post #42

Earlier quoted context omitted.

Yeah, that's true. GPU flops are really impressive in fp16, and more recently fp8 and fp4. When the 40xx GPUs came out, Tim Dettmers had a really cool blog looking into the numbers, and for fp8 one 4090 GPU had enough flops to match the best supercomputer somewhere in the 2000s, with an 8x 4090 build being top for a few years more. It's insane. But it gets nowhere close on fp64, where most of the physics simulations…

Other than weather simulations and nuclear explosions, what other supercomputer tasks are out there?

Although probably underfunded compared to the two that you listed, cosmology is probably up there as well. Things like early universe and galaxy formation, megastructures, etc.

There's likely also some need for fusion plasma containment and other related simulations.

Re: TOP500 at ISC’26: We have a New Number 1 Supercomputer

#74

Post by saagarjha »

adrian_b wrote: post #56
b33f wrote: post #30

Why are they not using GPUs? is it use cases that don't suit GPUs or because of the limitations they are imposing on themselves to use SMIC domestic chips?

If you can avoid GPUs, that is preferable. The reason is that with GPUs it is far more difficult to reach a great percentage of the maximum theoretical throughput. Most GPU programs reach only a very small fraction of what is theoretically possible, and in the best cases one may reach something like 50% to 60% of the maximum. This CPU-based supercomputer has demonstrated reaching 80% of the theoretical maximum throug…

> in the best cases one may reach something like 50% to 60% of the maximum

If all you need to do is matmuls then you can definitely go past this

Re: TOP500 at ISC’26: We have a New Number 1 Supercomputer

#75

Post by adrian_b »

saagarjha wrote: post #74
adrian_b wrote: post #56

Earlier quoted context omitted.

If you can avoid GPUs, that is preferable. The reason is that with GPUs it is far more difficult to reach a great percentage of the maximum theoretical throughput. Most GPU programs reach only a very small fraction of what is theoretically possible, and in the best cases one may reach something like 50% to 60% of the maximum. This CPU-based supercomputer has demonstrated reaching 80% of the theoretical maximum throug…

> in the best cases one may reach something like 50% to 60% of the maximum If all you need to do is matmuls then you can definitely go past this

You can go past this only on a matmul benchmark, which is seldom useful per se.

Linpack consists mostly of matmuls, but nonetheless there are additional operations that prevent GPUs to reach the high utilization of over 80% that is normal for CPUs, so that a throughput over 50% is considered good at the scale of supercomputers.

At the scale of a supercomputer, the utilization factor is considerably less than for an individual GPU or CPU, because the big matrix is split in blocks and the matrix multiplications are computed on different boards and in different racks, then the results are assembled, so there is a communication overhead.

The former Intel Xeon Phi, with a large number of cores that were weak except for their vector execution units, resembled GPUs in failing to reach a high utilization on Linpack.

Re: TOP500 at ISC’26: We have a New Number 1 Supercomputer

#76

Post by jamiejquinn »

fragmede wrote: post #71
NitpickLawyer wrote: post #42

Earlier quoted context omitted.

Yeah, that's true. GPU flops are really impressive in fp16, and more recently fp8 and fp4. When the 40xx GPUs came out, Tim Dettmers had a really cool blog looking into the numbers, and for fp8 one 4090 GPU had enough flops to match the best supercomputer somewhere in the 2000s, with an 8x 4090 build being top for a few years more. It's insane. But it gets nowhere close on fp64, where most of the physics simulations…

Other than weather simulations and nuclear explosions, what other supercomputer tasks are out there?

I was surprised to find out that material science uses up a significant amount of the compute time on the big research machines, at least here in the UK. It's both potentially lucrative research and computationally intensive. Combustion is the other big one that hasn't been mentioned but that's... Controversial.

Re: TOP500 at ISC’26: We have a New Number 1 Supercomputer

#78

Post by rbanffy »

microgpt wrote: post #69
antonvs wrote: post #40

Earlier quoted context omitted.

GPUs were for graphics. Now, they're mainly used for machine learning training and inference. The big tech companies are spending eye-watering amounts for GPUs - hundreds of billions of dollars a year each. That's the reason that Nvidia's market cap is at $4.66 trillion.

In between those, they were for Ethereum mining.

Before Ethereum, I believe Bitcoin sold a lot of GPUs as well.

Re: TOP500 at ISC’26: We have a New Number 1 Supercomputer

#80

Post by rbanffy »

fragmede wrote: post #71
NitpickLawyer wrote: post #42

Earlier quoted context omitted.

Yeah, that's true. GPU flops are really impressive in fp16, and more recently fp8 and fp4. When the 40xx GPUs came out, Tim Dettmers had a really cool blog looking into the numbers, and for fp8 one 4090 GPU had enough flops to match the best supercomputer somewhere in the 2000s, with an 8x 4090 build being top for a few years more. It's insane. But it gets nowhere close on fp64, where most of the physics simulations…

Other than weather simulations and nuclear explosions, what other supercomputer tasks are out there?

Weather modeling? Physics?
Post reply on HN