Earlier quoted context omitted.
> It ain't March 8 either I didn’t even notice that. That obviously means George is living in the future.
Or in Hong Kong.
it's the same picture.jpg
AMD YOLO
51–60 of 85 posts
Re: AMD YOLO
#52I would really love it if people on Hacker News could weigh in on how much of a moat they think CUDA really is. As in: How hard is it to use something else? If you started a project today how much would you want to get paid to not use CUDA? A lot of readers on this site have a good insight into this and it is a key question financial people are asking without the knowledge many people here possess.
> could weigh in on how much of a moat they think CUDA really is.
There's movement to implement CUDA libraries that work on non-Nvidia cards, but I guess adoption could be hindered by legal fears.
Re: AMD YOLO
#53> Bought in for a quarter million. Long term. It can always dip short term, but check back in 5 years. So basically he got 2 MI300s and is currently trying to pump AMD?
> 2 MI300s Not two cards, two boxes (with 8 cards each I'd assume).
It weighed 750 pounds so I think he got more than 16 cards. Or maybe they sent entire servers?
Re: AMD YOLO
#54AMD is probably undervalued, but Nvidia is clearly way overvalued. I don't think the reckoning will come from AMD stealing Nvidia's market share, it'll come when the hype bubble collapses and businesses start treating neural networks like commodities, running whatever is cheapest instead of the absolute most powerful. AMD is in a great position, because they make both great GPU/NPU hardware and great CPU hardware.
Is it possible someone will write a CUDA to AMD/Tensor/whatever transpiler (high-level emulator? I'm not sure of the right term) I thought there were a remarkably small number of ops that GPUs perform. Seems like a very high premium to pay for not wanting to rewrite in JAX or whatever.
Re: AMD YOLO
#55AMD is probably undervalued, but Nvidia is clearly way overvalued. I don't think the reckoning will come from AMD stealing Nvidia's market share, it'll come when the hype bubble collapses and businesses start treating neural networks like commodities, running whatever is cheapest instead of the absolute most powerful. AMD is in a great position, because they make both great GPU/NPU hardware and great CPU hardware.
Is it possible someone will write a CUDA to AMD/Tensor/whatever transpiler (high-level emulator? I'm not sure of the right term) I thought there were a remarkably small number of ops that GPUs perform. Seems like a very high premium to pay for not wanting to rewrite in JAX or whatever.
AMD hired 1 guy to do it and then fired him
Re: AMD YOLO
#56I would really love it if people on Hacker News could weigh in on how much of a moat they think CUDA really is. As in: How hard is it to use something else? If you started a project today how much would you want to get paid to not use CUDA? A lot of readers on this site have a good insight into this and it is a key question financial people are asking without the knowledge many people here possess.
Whenever a new AI model gets released and is available for the public. From the last few I've tried they were always NVIDIA only because I assume that's what the researchers had at their disposal.
Re: AMD YOLO
#57I saw a video the other day that showed a new AMD laptop processor that is comparable to the Apple processors in performance and battery life. That was very surprising and also a great thing for windows or Linux laptops. But at the same time, the market for these and the potential for profit isn’t really that big. Consumers are willing to pay for a premium for Apple but not anyone else.
Do you have a link? I'd like to know which model that was.
Re: AMD YOLO
#58Earlier quoted context omitted.
> foreseeable future I think this is the arguable part. The more AI compute becomes valuable, the more reason there is to divest from their software moat. Their hardware is good, but not unassailable. I think their modest (by tech hype standards) P/E is recognition of this.
But they are making strides in robotics already. Jensen is super smart, business-savvy AND hard-working. These are some of many reasons I own NVDA.
>Jensen is super smart, business-savvy AND hard-working
Those traits are table stakes for running a Fortune 50 company. Before 2019 when the AI boom came out of nowhere, what was going on? Nvidia was an okay company, but not a real over-performer.
Re: AMD YOLO
#59Earlier quoted context omitted.
Is it possible someone will write a CUDA to AMD/Tensor/whatever transpiler (high-level emulator? I'm not sure of the right term) I thought there were a remarkably small number of ops that GPUs perform. Seems like a very high premium to pay for not wanting to rewrite in JAX or whatever.
AMD hired 1 guy to do it and then fired him
I wondered if John Carmack would take his strong knowledge of low level hardware and interest in AI and work on this, but he seems to be working on a non LLM flavor of AI.
Re: AMD YOLO
#60AMD is probably undervalued, but Nvidia is clearly way overvalued. I don't think the reckoning will come from AMD stealing Nvidia's market share, it'll come when the hype bubble collapses and businesses start treating neural networks like commodities, running whatever is cheapest instead of the absolute most powerful. AMD is in a great position, because they make both great GPU/NPU hardware and great CPU hardware.
Is it possible someone will write a CUDA to AMD/Tensor/whatever transpiler (high-level emulator? I'm not sure of the right term) I thought there were a remarkably small number of ops that GPUs perform. Seems like a very high premium to pay for not wanting to rewrite in JAX or whatever.
Lack of translation layer wasn't the problem, AMD's shit being completely broken (for this purpose) was the problem. Driver-level broken, possibly hardware-level broken. Black screens, restarts, wrong answers, threads full of people shouting into the void year after year about the same behavior before seeing the light and buying nvidia.
Then NVDA pumped by trillions and AMD did not and even AMD's crack team of trained denial specialists could no longer stay the course, so they started to turn the ship. But that was only a year or two ago and even the tiniest changes take years in hardware land so the biggest sins are still baked in to the latest chips, but at least the software has started to suck a bit less.
I no longer see 100% of the people who try to build on AMD compute run away screaming. Many do, but not all. That's a change, an important and positive one. If AMD keeps it up maybe they can save us from 80% margins on matrix multiplication after all.