MI300X vs. H100 vs. H200 Benchmark Part 1: Training – CUDA Moat Still Alive

Re:MI300X vs. H100 vs. H200 Benchmark Part 1: Training – CUD... by AlotOfReading on Sunday December 22, 2024 @08:45PM (#42490960)

Sadly common at hardware companies. The most extreme case I've heard of is ASML, who supposedly doesn't keep any machines of their own. They test against "almost-ready" machines right before they go out the door to customers.

[ Parent | on HN ]

Re:MI300X vs. H100 vs. H200 Benchmark Part 1: Training – CUD... by dmoy on Sunday December 22, 2024 @08:54PM (#42491008)

ASML might be an extreme outlier though, don't those things cost like $50 million+ each?

[ Parent | on HN ]

Re:MI300X vs. H100 vs. H200 Benchmark Part 1: Training – CUD... by tonetegeatinst on Sunday December 22, 2024 @09:27PM (#42491177)

Many for last gen process nodes, and from a second or third hand supplier if you could even find one. ASML makes very few fully working machines each year, and the cost and throughput those machines have is astronomical.

They have spare parts you'd bet, and I'd bet they have some SLA agreement with each customer where an engineer is basically on call nearby in case a single thing dosnt work or a random part breaks or needs servicing.

Asianometry did a great video on the cost of downtime when it comes to ASML device in any fab. While I am not directly in this field and can't speak to the accuracy of the numbers john gives, he does not seem one to just make stuff up as his quality of video production for niche topics is quite good.

[ Parent | on HN ]

Re:MI300X vs. H100 vs. H200 Benchmark Part 1: Training – CUD... by bobmcnamara on Sunday December 22, 2024 @11:04PM (#42491654)

Almost a decade ago KFAB had a fire, power was cut, everything in process was dumped, they planned to restart but ended up being was cheaper to close the whole facility .

Probably for the best though, KFAB had been discharging several tons of solvents, cleaning agents, and reagents per year into the surrounding area [for as long as it ran](https://enviro.epa.gov/facts/tri/ef-facilities/#/Release/640...)

https://enviro.epa.gov/facts/tri/ef-facilities/#/Release/640...

[ Parent | on HN ]

Re:MI300X vs. H100 vs. H200 Benchmark Part 1: Training – CUD... by adgjlsfhk1 on Sunday December 22, 2024 @10:52PM (#42491606)

try 400 mil

[ Parent | on HN ]

Re:MI300X vs. H100 vs. H200 Benchmark Part 1: Training – CUD... by amirhirsch on Sunday December 22, 2024 @09:02PM (#42491062)

Put a 1 or 2 in front of that

[ Parent | on HN ]

Re:MI300X vs. H100 vs. H200 Benchmark Part 1: Training – CUD... by rahkiin on Monday December 23, 2024 @03:12AM (#42492734)

That’s why they cooperate closely with imec and their FAB in Leuven

[ Parent | on HN ]

Re:MI300X vs. H100 vs. H200 Benchmark Part 1: Training – CUD... by whazor on Monday December 23, 2024 @03:54AM (#42492928)

Actually, they shipped not-ready machines to customers. In hope they could find solutions that can fix the machine later.

[ Parent | on HN ]

Re:MI300X vs. H100 vs. H200 Benchmark Part 1: Training – CUD... by Flenkno on Monday December 23, 2024 @08:39AM (#42494304)

You are comparing a machine the size of a container with GPUs?

its nice to be aware of this but this is so fastly different from a critisism point of view that i don't think that matters.

[ Parent | on HN ]

Re:MI300X vs. H100 vs. H200 Benchmark Part 1: Training – CUD... by AlotOfReading on Monday December 23, 2024 @09:41AM (#42494696)

It's not a criticism, it's an extreme example from a company people know that I have no particular NDA restrictions with.

[ Parent | on HN ]

Re:MI300X vs. H100 vs. H200 Benchmark Part 1: Training – CUD... by llamaimperative on Monday December 23, 2024 @12:09PM (#42495880)

I found it interesting ¯\_(ツ)_/¯

[ Parent | on HN ]