Not that this really takes away from the substance of the article, but the first two paragraphs are giving heavy Claude smell. Semicolons, em dashes, "That sequencing matters"... I guess I'm just a little surprised that anyone could be arsed to take on a hardware project like this but can't be arsed to write their own introduction.
> 4× RTX PRO 6000 Blackwell Workstation (GB202, 96 GB GDDR7, 600 W)<p>Those are SM120 so no tmem/tcgen05 and lack of support in main libraries (it's like everybody is focusing on B300/SM100).<p>For that money I'd buy a single B300, similar total AI TOPS, similar GPU bandwidth aggregated, and only 25% less total memory (probably saved in less implementation complexity), half the energy consumption...<p>Also by having all SMs local they have the special L1-level interconnect. SMs can collaborate on the same GEMM. And a bunch of other nice features.<p>Or, you know, rent it.
Pass the generated text through some kind of quality prompt. It’s got too much filler right now.<p>The story is interesting but it’s hard to read because it’s hard to tell which parts are meaningful and which parts are filler.<p>E.g. “we pulled the card cold - straight from the rig to the workbench”. Okay, but why would going straight from the rig to the workbench make it cold? If anything it would be warm. But it turns out the temperature is meaningful in your story.
What a time to be alive, I remember 10 years ago as a poor student waiting to buy a ATI Radeon X1600 Pro with 256mb, yes 256mb of RAM.<p>It cost about £190 in 2006.<p>Now we have GPUs that are in tens of thousands of pounds with insane performance, but what would their price be without the AI and Datacentre squeeze?
Ok, how are people powering these things? 2.4kW is well beyond a standard circuit in the US. Are people having 240V/30A circuits installed? Are they hijacking the dryer plugs? EV charger plugs? Hottub circuits?
240V-20A circuits will handle 3.8kW continuous. It’s probably a 240V-20A circuit, as that is what the power supplies typically want. Also, easy to convert an outlet to 240V, if the breaker is dedicated to that outlet. Just requires swapping the breaker and the outlet, not the wires.
Chaining two PSUs on separate circuits is also an option. If they're using the MaxQ versions though, the total GPU power draw is only ~1200W. The bigger question to me is how are they cooling it? Sticking an AC in that room just doubles the power draw issues.
It is basically on 2 different circuits/breakers. Asus wrx90e supports 2 psu as well. You may need to synchronize both psu and several adapter for this is available in Amazon. Soon planning to upgrade it to 240V
exactly, I had a 220v 30a circuit installed to run a multi-GPU server in my basement.<p>I'm air cooling so I set -pl 450 so I'm not running them all at the full 600w
I wonder whether those cards ran the model that wrote the nonsense about the forces involved.<p>Hint: when you have a piece of metal stuck with thermal goop to a lot of components, the force doesn’t “concentrate” on one of them. You need to detach it from each one with however much force is needed to detach it from <i>that component</i>.
It's difficult to speculate as to the exact failure from blurry pictures but the solder on that choke (inductor) looks terrible.<p>Something went wrong in manufacturing. The solder should have wicked to cover the entire pad, not just a small square, and there should be no (brown) discoloration.
Cool post. FYI you might be better off getting one big fan for your "radiator" instead of lots of little fans. Big fans don't need to spin as fast as small fans to push the same amount of air. So they run a lot quieter.
Out of curiosity, what are you training with these cards?
Is that little computer training LLMs from scratch all by itself? That must take years to get any kind of progress, given the scale of training other providers do. Where do you get the training data from?
Most of the training i am working on is with post training. You can do so much with a system that is running 24/7
You can train TinyStories in a few hours on retail hardware, and this is a highly illuminating experience that I can recommend for everyone.
[dead]
Complete side note, but I can’t work out how the author managed to mistype “at” as “Δt”.<p>Edit: reading fail on my part, nothing to see here.
I caught that too, probably a qwen bug (;
huh? the only Δt in the article is used correctly.<p>> With 18× 140 mm of surface, the fans run quietly and the coolant Δt across the rads stays small
Converting four RTX PRO 6000 Blackwell cards to waterblocks, finding a VRM choke loose on the workbench, and getting back to 41k tok/s.
If you want ready, well engineered, water-cooled multi-GPU research workstations, my colleagues at <a href="https://comino.com" rel="nofollow">https://comino.com</a> build and sell them. Or you can purchase fitted waterblocks from them for many GPUs, and build your own.
Ditch the tiny DC fans. Build a shroud and switch to a single ac-powered industrial blower / duct fan.
AI slop post.
Why does this post sound like it's an AI story based on the inputs from the engineer?<p>The phrasing is very claude like:<p>"That cracked joint is the whole story. The card had passed initial bring-up and ran fine at light loads for a week."<p>"That sequencing matters — it’s why we have a story to tell. The pilot card failed, taught us a lesson, and the lesson is the reason the other three went on without incident."<p>"Driver swaps, CUDA reinstalls, and inference-engine theories were dead ends I spent hours on. The failure pattern itself told the story — listen to it earlier."