Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AMD may get across the CUDA moat

hpcwire.com

Post reply on HN
Print view View original

31–40 of 312 posts

Re: AMD may get across the CUDA moat

#31

Post by throwitawayfam »

IntelMiner wrote: post #27
mattnewton wrote: post #4

Earlier quoted context omitted.

Re: startups, Geohotz raised a few million for this already. https://tinygrad.org/

Didn't he do what he always does. Rake in a ton of money, fart around and then cash out exclaiming it's everyone else's fault? The way he stole Fail0verflow's work with the PS3 security leak after failing to find a hypervisor exploit for months absolutely soured any respect I had for him at the time

Yep, did exactly that. IMO he threw a fit, even though AMD was working with him squashing bugs. https://github.com/RadeonOpenCompute/ROCm/issues/2198#issuec...

Re: AMD may get across the CUDA moat

#32

Post by AnthonyMouse »

sumtechguy wrote: post #20
klysm wrote: post #7

Earlier quoted context omitted.

A simplistic economic take would suggest that the competition would result in lower prices, but given two players in the market who knows.

It is oligopoly pricing. https://www.investopedia.com/terms/o/oligopoly.asp With that few competitors pricing would not change much.

That's mostly when there isn't a lot of price elasticity of demand. If you're Comcast and Verizon, each customer wants one internet connection and you're not going to change the size of the market much by offering better prices.

If you're AMD and NVIDIA and lowering the price would double the number of customers, you might very well want to do that, unless you're supply constrained -- which has been the issue because they're both bidding against everyone else for limited fab capacity. But that should be temporary.

This is also a market with a network effect. If all your GPUs are $1000 and nobody can afford them then nobody is going to write code for them, and then who wants them? So the winning strategy is actually to make sure that there are kind of okay GPUs available for less than $300 and make sure lots of people have them, then sell very expensive ones that use the same architecture but are faster.

That has been the traditional model, but the lack of production capacity meant that they've only been making the overpriced ones recently. Which isn't actually in their interests once the supply of fab capacity loosens up.

Re: AMD may get across the CUDA moat

#33

Post by ad404b8a372f2b9 »

sumtechguy wrote: post #20
klysm wrote: post #7

Earlier quoted context omitted.

A simplistic economic take would suggest that the competition would result in lower prices, but given two players in the market who knows.

It is oligopoly pricing. https://www.investopedia.com/terms/o/oligopoly.asp With that few competitors pricing would not change much.

Prices seemed to have lowered when AMD came out with CPUs competitive with Intel's.

Re: AMD may get across the CUDA moat

#34

Post by ilc »

wmf wrote: post #25
rafaelmn wrote: post #17

Earlier quoted context omitted.

If the margins and demand is there Intel will eventually show up

Intel already showed up three or four times but their software is as bad as AMD's used to be.

Thankfully, software can be fixed over time as AMD has shown. Lack of another competitor can't be fixed as easily.

Re: AMD may get across the CUDA moat

#35

Post by smoldesu »

javchz wrote: post #26

CUDA is the only reason I have an Nvidia card, but if more projects start migrating to a more agnostic environment, I'll be really grateful. Running Nvidia in Linux isn't as much fun. Fedora and Debian can be incredibly reliable systems, but when you add an Nvidia card, I feel like I am back in Windows Vista with kernel crashes from time to time.

Those problems might just be GNOME-related at this point. I've been daily-driving two different Nvidia cards for ~3 years now (1050 Ti then 3070 Ti) and Wayland has felt pretty stable for the past 12 months. The worst problem I had experienced in that time was Electron and Java apps drawing incorrectly in xWayland, but both of those are fixed upstream.

I'm definitely not against better hardware support for AI, but I think your problems are more GNOME's fault than Nvidia's. KDE's Wayland session is almost flawless on Nvidia nowadays.

Re: AMD may get across the CUDA moat

#36

Post by arsome »

smoldesu wrote: post #35
javchz wrote: post #26

CUDA is the only reason I have an Nvidia card, but if more projects start migrating to a more agnostic environment, I'll be really grateful. Running Nvidia in Linux isn't as much fun. Fedora and Debian can be incredibly reliable systems, but when you add an Nvidia card, I feel like I am back in Windows Vista with kernel crashes from time to time.

Those problems might just be GNOME-related at this point. I've been daily-driving two different Nvidia cards for ~3 years now (1050 Ti then 3070 Ti) and Wayland has felt pretty stable for the past 12 months. The worst problem I had experienced in that time was Electron and Java apps drawing incorrectly in xWayland, but both of those are fixed upstream. I'm definitely not against better hardware support for AI, but I…

If GNOME can tank the kernel, it ain't GNOME's fault.

Re: AMD may get across the CUDA moat

#38

Post by AnthonyMouse »

dylan604 wrote: post #29
Havoc wrote: post #22

Earlier quoted context omitted.

>They are just too late even if they catch up. Late certainly, too late I don't think so. If you can field a competitively priced consumer card that can run llama fast then you're already halfway there because then the ecosystem takes off. Especially since nvidia is being really stingy with their vram amounts. H100 & datacenter is a separate battle certainly, but on mindshare I think some deft moves from AMD will get…

>If you can field a competitively priced consumer card if this unicorn were to show up, what's to say that all the non-consumers won't just scarf up these equally performant yet lower priced cards causing the supply-demand situation we're in now? the only difference would be a sudden supply of the expensive Nvidia cards that nobody wants because of their price.

The thing that causes it to be competitively priced is having enough production capacity to prevent that from happening.

One way to do that may be to produce a card on an older process node (or the existing one when a new one comes out) that has a lot of VRAM. There is less demand for the older node so they can produce more of them and thereby sell them for a lower price without running out.

Re: AMD may get across the CUDA moat

#39

Post by mnau »

pixelesque wrote: post #15

Does AMD have a solution to forward device combatibility (like PTX for NVidia)? Last time I looked into ROCm (two years ago?), you seemed to have to compile stuff explicitly for the architecture you were using, so if a new card came out, you couldn't use it without a recompile.

Not natively, but AdaptiveCpp (previously hiSycl, then OpenSycl) has a single source single compiler pass, where they basically store LLVM IR as an intermediate representation.

https://github.com/AdaptiveCpp/AdaptiveCpp/blob/develop/doc/...

Performance penalty was within ew precents, at least according to the paper (figure 9 and 10) https://cdrdv2-public.intel.com/786536/Heidelberg_IWOCL__SYC...

Re: AMD may get across the CUDA moat

#40

Post by quitit »

binarymax wrote: post #6

And the question for most that remains once AMD catches up: will the duopoly result in lower prices to a reasonable level for hobbyists or bootstrapped startups, or will AMD just gouge like NVidia?

I think in this case the changes needed to make AMD useful will open the market to other players as well (e.g. Intel).

PyTorch is already walking down this path and while CUDA-based performance is significantly better, that is changing and of course an area of continued focus.

It's not that people don't like Nvidia, rather it's just that there is a lot of hardware out there that can technically perform competitively, but the work needs to be done to bring it into the circle.

Post reply on HN