4 ms·
AMD Launches Instinct MI300X AI Accelerator, Up to 60% Faster Than Nvidia H100
- filterfiber 3y agoI really hope AMD gains more ground on the software side. I know they are "technically supported" by onyx/pytorch etc. But almost every project I've seen there's a bunch of asterisks, a bunch of people needing different work arounds, and subpar performance compared to the hardware. Maybe it's because most people have their consumer cards? Nvidia however doesn't have this issue, people on pascal are only just now having compatibility issues with not supporting a high enough cuda version. The dream is they release a 4090-type competitor for a few grand, it doesn't even need full 4090-level performance, just strap 48gb+ of HBM and it would be a absolutely incredible deal for individuals learning/researching -> who then use the full enterprise stuff for full training/commercial inference.
- rreichman 3y agoTheir software is definitely improving and will get a whole lot better in the next few years. It's worth many billions for companies to exit the Nvidia lock-in.
- fariszr 3y agoAs long, as it doesn't support normal consumer hardware, AMD won't get the support Nvidia GPUs enjoy from community projects.
- sbininit 3y agoI suspect that those who would otherwise be buying thousands of H100s would simply put in the dev work to make the software function well on AMD hardware
- idonotknowwhy 3y agoI'm hoping for the same thing, but with Intel ARC. Intel gpus are cheaper, but have even worse ML software for consumer projects
- frognumber 3y agoI'm optimistic. * Intel has a history of getting this type of software working. * AMD doesn't. I don't think I'll ever trust AMD after... experiences, but Intel I'm bullish on Intel. Arc A770 seems like by far the bang for the buck once software lands. 16GB for $300 if you find a good deal. Decent performance. Stick four of them in a system, and you've got 64GB for $1200 and a pretty decent amount of compute. Buy five systems like that, and you can run the largest open-source LLMs out there with Alpa for ~$6K. That brings Facebook's GPT3-grade OPT models from corporate down to serious enthusiast level.
- rreichman 3y agoAccording to this and SemiAnalysis[1], the MI300X seems impressive. Better than Nvidia's H100 and maybe a bit inferior to the H200. Why is Nvidia worth $1 trillion more? [1] - https://www.semianalysis.com/p/amd-mi300-performance-faster-than https://www.semianalysis.com/p/amd-mi300-performance-faster-...
- mattnewton 3y agoCUDA
- rreichman 3y agoCUDA is becoming less important over time
- frognumber 3y agoI'm not sure it is. You can do more with non-CUDA now than ever before. At the same time, the number of things you can do only with CUDA is greater than ever before too. The whole field it growing, and it's hard to compare.
- tedunangst 3y ago$9B in profit vs $300M will have that effect.
- rreichman 3y agoSure but the future is not the past
- riku_iki 3y agoAMD p/e is 900 vs 60 for NVDA, so investors already heavily invested into AMD future.
- tedunangst 3y agoTime to buy those calls.
- Racing0461 3y agoWe need competition in the ai gpu space.