5 ms·

Why are AMD or any other competition (I don't follow this sector very closely) not taking advantage of this and compete on price?

metaphor 3y ago
The big picture I'm seeing is there's a game being played[1], but to your point, AMD did respond to RTX 4070 launch by dropping prices on RX 6800 series GPUs[2]. [1] https://scholarworks.umass.edu/econ_workingpaper/343/ https://scholarworks.umass.edu/econ_workingpaper/343/ [2] https://news.ycombinator.com/item?id=35663511 https://news.ycombinator.com/item?id=35663511
epolanski 3y ago
Because gaming GPUs are not really profitable to AMD. As a reference, between 2008 and 2011 when ATI/AMD had the biggest technological lead over Nvidia's disaster architectures, even then when AMD was at near 50% market share it merely posted 19M of profits across 3 financial years! In the same time frame, despite terrible performances, Fermi architecture already going into the 400w consumption range, even then Nvidia posted billions of profits across the same time frame. AMD has simply stopped competing with Nvidia on prices, it's nonsense to them. Nowadays both Nvidia and AMD sell more GPUs to datacenters and super computers, where net margins are also twice as high than they do to gamers. Essentially both AMD and Nvidia are abandoning the gaming aspect of their hardware, stopped competing price-wise, instead they prefer both to sit on higher margins in the gaming market and focus on GPGPU aspect for business, research, automotive, AI, etc, etc, where the real money and growth is. The fact that this hardware is also able to crunch rendering calculations is more and more of a side effect.
- hoseja 3y ago
  We need games that utilize local neural nets.
  - mepian 3y ago
    It seems that DLSS/XeSS is the only feasible application of local neural nets to games right now, everything else is too compute-hungry.
- pdpi 3y ago
  AMD has notably cornered the console market with the Xbox One and Series X|S, PS4 and PS5. With the Steam Deck and now the Asus ROG Ally, they're starting to get themselves a position in the handheld market too.
  - themitigating 3y ago
    Is OEM integration profitable though? I would imagine Microsoft and Sony squeezed the best deal and maybe it will bring profit in later years when the research debt is paid
    - soganess 3y ago
      Judging by their conservative clocks and odd configurations (ex steam deck is a 4-core) these are some of the lower bind parts that would otherwise not be high-value final products on the open PC market. Perhaps even "defective" (ex failed PCIe lanes) in some way that makes them useless. I always figured this was a "use the whole pig" situation.
  - epolanski 3y ago
    Sure, but those aren't a company-defining revenues or where huge growth is.
- prettyStandard 3y ago
  This isn't entirely make sense to me, I thought AMD was very far behind on GPGPU. Do these supercomputer and data center projects use things like ROCm and OpenCL, and solve the software issues on their side? Does AMD compete on price there?
  - epolanski 3y ago
    The world's fastest supercomputer, Frontier at Oak Ridge, and the third fastest Lumi, in Finland are both exclusively made with AMD hardware. https://en.wikipedia.org/wiki/Frontier_(supercomputer) https://en.wikipedia.org/wiki/Frontier_(supercomputer) https://en.wikipedia.org/wiki/LUMI https://en.wikipedia.org/wiki/LUMI There's lots of demands for AMD hardware in datacenters and supercomputers.
    - prettyStandard 3y ago
      I didn't say it wasn't true I was just trying to say it didn't make sense to me. If you're trying to build a giant supercomputer it seems like you would pick something where the software works. But maybe they are using ROCm. It seems like that only works for the higher end stuff. Maybe they're planning on running stuff that's already been built on OpenCL, and AMD is competing on the price for the publicity.
meltyness 3y ago
Intel doesn't really have a competing product. AMD and Nvidia sell the same thing made at the same place in Taiwan. Nvidia dove in like 1998. AMD sold it's moat in like, 2005.
reissbaker 3y ago
Contrary to what other posters are saying, I do think AMD is actually trying to compete; for example, they finally gave up on OpenCL and are trying to play catch-up to Nvidia via ROCm, effectively allowing you to run CUDA on AMD's server GPU offerings. And on the gaming side their top-of-the-line offering, the 7900 XTX, is ~$600 cheaper (aka 60%!) at MSRP than Nvidia's top-of-the-line RTX 4090 — and the prices you'd pay from e.g. resellers on eBay are even more skewed towards Nvidia. That being said, they're just way, way behind. Nvidia can charge what they do simply because AMD isn't great at making GPUs. In gaming AMD never caught up to Nvidia's raytracing perf, and AMD's attempt to clone DLSS is significantly less performant. Meanwhile in AI/ML the industry consolidated around CUDA because OpenCL was never that great, and — Nvidia owning the standard, and having built around it for over a decade, gives them a really big advantage. Plus, every GPU sold in the last decade+ by Nvidia supports CUDA, whereas only the new server GPUs from AMD support ROCm. Nvidia has executed incredibly well for decades: they pioneered shader programming, CUDA, raytracing, and DLSS. Arguably they even invented the GPU itself with the GeForce 256 in 1999 (as opposed to earlier "accelerator" cards). I struggle to think of anything AMD has done in the GPU space that matches even one of those accomplishments... Their only slight win is having OSS drivers for Linux, but honestly, that doesn't sell a whole lot of GPUs. AMD caught up to Intel on the CPU side in part thanks to Intel's failures. Nvidia just hasn't fumbled. Edit: Also, Intel is trying to edge their way into the GPU market as well with their Arc line of discrete GPUs, which is relatively affordable. They compete well on price, especially at the lower end: the A750 and A380 obliterate anything at their price points ($239 and $119, respectively). But, well, "at their price points" is a pretty critical part of that sentence. Their top-tier card, the A770, can barely eke out wins against the lower end of Nvidia's last-gen cards. And re: 2008-2011 ATI comparisons from a sibling comment — by 2010 Nvidia's Fermi architecture was reliably beating ATI again, albeit with higher power draw, and in 2007 they had just launched CUDA which ended up being the most lucrative market segment for GPUs. In retrospect I think Fermi was much less of a flop than it seemed (and it was a fairly brief one compared to Intel's CPU woes).
- nl 3y ago
  ROCm development is understaffed and continually considerably behind CUDA too.

metaphor 3y ago

The big picture I'm seeing is there's a game being played[1], but to your point, AMD did respond to RTX 4070 launch by dropping prices on RX 6800 series GPUs[2]. [1] https://scholarworks.umass.edu/econ_workingpaper/343/ https://scholarworks.umass.edu/econ_workingpaper/343/ [2] https://news.ycombinator.com/item?id=35663511 https://news.ycombinator.com/item?id=35663511

epolanski 3y ago

Because gaming GPUs are not really profitable to AMD. As a reference, between 2008 and 2011 when ATI/AMD had the biggest technological lead over Nvidia's disaster architectures, even then when AMD was at near 50% market share it merely posted 19M of profits across 3 financial years! In the same time frame, despite terrible performances, Fermi architecture already going into the 400w consumption range, even then Nvidia posted billions of profits across the same time frame. AMD has simply stopped competing with Nvidia on prices, it's nonsense to them. Nowadays both Nvidia and AMD sell more GPUs to datacenters and super computers, where net margins are also twice as high than they do to gamers. Essentially both AMD and Nvidia are abandoning the gaming aspect of their hardware, stopped competing price-wise, instead they prefer both to sit on higher margins in the gaming market and focus on GPGPU aspect for business, research, automotive, AI, etc, etc, where the real money and growth is. The fact that this hardware is also able to crunch rendering calculations is more and more of a side effect.

hoseja 3y ago
We need games that utilize local neural nets.
- mepian 3y ago
  It seems that DLSS/XeSS is the only feasible application of local neural nets to games right now, everything else is too compute-hungry.
pdpi 3y ago
AMD has notably cornered the console market with the Xbox One and Series X|S, PS4 and PS5. With the Steam Deck and now the Asus ROG Ally, they're starting to get themselves a position in the handheld market too.
- themitigating 3y ago
  Is OEM integration profitable though? I would imagine Microsoft and Sony squeezed the best deal and maybe it will bring profit in later years when the research debt is paid
  - soganess 3y ago
    Judging by their conservative clocks and odd configurations (ex steam deck is a 4-core) these are some of the lower bind parts that would otherwise not be high-value final products on the open PC market. Perhaps even "defective" (ex failed PCIe lanes) in some way that makes them useless. I always figured this was a "use the whole pig" situation.
- epolanski 3y ago
  Sure, but those aren't a company-defining revenues or where huge growth is.
prettyStandard 3y ago
This isn't entirely make sense to me, I thought AMD was very far behind on GPGPU. Do these supercomputer and data center projects use things like ROCm and OpenCL, and solve the software issues on their side? Does AMD compete on price there?
- epolanski 3y ago
  The world's fastest supercomputer, Frontier at Oak Ridge, and the third fastest Lumi, in Finland are both exclusively made with AMD hardware. https://en.wikipedia.org/wiki/Frontier_(supercomputer) https://en.wikipedia.org/wiki/Frontier_(supercomputer) https://en.wikipedia.org/wiki/LUMI https://en.wikipedia.org/wiki/LUMI There's lots of demands for AMD hardware in datacenters and supercomputers.
  - prettyStandard 3y ago
    I didn't say it wasn't true I was just trying to say it didn't make sense to me. If you're trying to build a giant supercomputer it seems like you would pick something where the software works. But maybe they are using ROCm. It seems like that only works for the higher end stuff. Maybe they're planning on running stuff that's already been built on OpenCL, and AMD is competing on the price for the publicity.

meltyness 3y ago

Intel doesn't really have a competing product. AMD and Nvidia sell the same thing made at the same place in Taiwan. Nvidia dove in like 1998. AMD sold it's moat in like, 2005.

reissbaker 3y ago

Contrary to what other posters are saying, I do think AMD is actually trying to compete; for example, they finally gave up on OpenCL and are trying to play catch-up to Nvidia via ROCm, effectively allowing you to run CUDA on AMD's server GPU offerings. And on the gaming side their top-of-the-line offering, the 7900 XTX, is ~$600 cheaper (aka 60%!) at MSRP than Nvidia's top-of-the-line RTX 4090 — and the prices you'd pay from e.g. resellers on eBay are even more skewed towards Nvidia. That being said, they're just way, way behind. Nvidia can charge what they do simply because AMD isn't great at making GPUs. In gaming AMD never caught up to Nvidia's raytracing perf, and AMD's attempt to clone DLSS is significantly less performant. Meanwhile in AI/ML the industry consolidated around CUDA because OpenCL was never that great, and — Nvidia owning the standard, and having built around it for over a decade, gives them a really big advantage. Plus, every GPU sold in the last decade+ by Nvidia supports CUDA, whereas only the new server GPUs from AMD support ROCm. Nvidia has executed incredibly well for decades: they pioneered shader programming, CUDA, raytracing, and DLSS. Arguably they even invented the GPU itself with the GeForce 256 in 1999 (as opposed to earlier "accelerator" cards). I struggle to think of anything AMD has done in the GPU space that matches even one of those accomplishments... Their only slight win is having OSS drivers for Linux, but honestly, that doesn't sell a whole lot of GPUs. AMD caught up to Intel on the CPU side in part thanks to Intel's failures. Nvidia just hasn't fumbled. Edit: Also, Intel is trying to edge their way into the GPU market as well with their Arc line of discrete GPUs, which is relatively affordable. They compete well on price, especially at the lower end: the A750 and A380 obliterate anything at their price points ($239 and $119, respectively). But, well, "at their price points" is a pretty critical part of that sentence. Their top-tier card, the A770, can barely eke out wins against the lower end of Nvidia's last-gen cards. And re: 2008-2011 ATI comparisons from a sibling comment — by 2010 Nvidia's Fermi architecture was reliably beating ATI again, albeit with higher power draw, and in 2007 they had just launched CUDA which ended up being the most lucrative market segment for GPUs. In retrospect I think Fermi was much less of a flop than it seemed (and it was a fairly brief one compared to Intel's CPU woes).

nl 3y ago
ROCm development is understaffed and continually considerably behind CUDA too.