I'm an amateur, but I have code that I think could probably dispatch threads pretty efficiently on the Cheyenne thru it's management system simply because it's all xeons distributed. If I can run it on my personal 80-core cluster, I could get it to run on Cheyenne back then.
But hitting the roofline on those AMD GPGPU's? I'd probably get nowhere fucking close.
That is the thing that Cheyenne was built for. People doing CFD research with x86 code that was already nicely parallelized via OpenMPI or whathaveyou.
I used to build small clusters and use supercomputers and I can't imagine it's fun to build a super computer. It requires a massive infrastructure and significant employee base, and individual component failures can take down entire jobs. Finding enough jobs to keep the system loaded 24/7 while also keeping the interconnect (which was 15-20% of the total system cost) busy, and finding the folks who can write such jobs, is not easy. Even then, other systems will be constantly nipping at your heels with newer/cheaper/smaller/faster/cooler hardware.
Thanks for the feedback. You make a lot of good points. I've built a 150,000 GPU system previously, but it was lower end hardware. It was a lot of fun to make it run smoothly with its own challenges.
It doesn't take a lot of employee's, we did the above on essentially two technical people. Those same two are working on this business.
Finding workloads/jobs is definitely going to be an interesting adventure, that said, the need for compute isn't going away. By offering hard to get hardware at reasonable rates and contract lengths, I believe we are in a good position on that front, but time will tell.
We are only buying the best of the best that we can get today. The plan is to continuously cycle out older hardware as well as not pick sides on one over another. This should help us keep pace with other systems.
150K GPU with two people... presumably, 8 GPU/host, you had close to 20K servers.
I can't really see how that's achievable with only two people, given the time to install hardware, maintain it, deal with outages and planned maintainence and testing, etc. Note: I worked at Google and interfaced with hwops so I have some real-world experience to compare to.
Building a 150K GPU system without a well-understood customer base seems a bit crazy to me. You will either become a hyperscale, serve a niche, or go out of business, I fear.
The ASRock BC-250's we deployed were 12 individual blades and those were all PXE booted. We deployed 20,000 of those blades across 2 data centers. This was a massive feat of engineering, especially during covid where I couldn't even access the machine directly. Built a whole dashboard to monitor it all too.
I know, I can't believe we did it either, but we did. Software automation was king. I built a single binary that ran on each individual host and knew how to self configure / optimize everything. Idempotently. Even distributing upgrades to the binary was a neat challenge that I solved perfectly, in very creative ways.
Today, we are starting much smaller. Literally from zero/scratch. Given the cost of MI300x, I doubt we will ever get to 150k GPUs, that's an absurd amount of money, but who knows.
But who did the wiring? Even with blades which consolidate much of the cabling, there's still a tremendous amount of work to build the interconnect. On typical large systems I've seen a small team 3-5 guys working weeks+ to wire a modest DC.
We'd hire the initial deployment out to temporary contractors. It just took a few weeks to get a large deployment out. The hard part was the 12 GPUs needed to be inserted at the DC, which took a bunch of effort. Once it was done we generally had 1-2 people on the ground in the data centers to deal with breakfixes. Either contractors or supplied by the DC.
For this venture, again, we are starting small, so we are just flying to the DC and doing it ourselves. There are also staff there that are technical enough to swap stuff out when we need it. The plan will be to just hire one of their staff as our own.
I don't think we will make it for this next deployment due to time constraints, but ideally in our near future, we will go full L11. Assemble and ship out full racks at the manufacturer/VAR, bolt em down, wire them up and ready to go. That is my dream... we will see if we get there. L11 is hard cause a single missing cable can hold up an entire shipment.
But hitting the roofline on those AMD GPGPU's? I'd probably get nowhere fucking close.
That is the thing that Cheyenne was built for. People doing CFD research with x86 code that was already nicely parallelized via OpenMPI or whathaveyou.