4 ms·

And I can't name several customers with very deep pockets working on their own chips to squeeze/compete with NVDA.

alephnerd 2y ago
> And I can't name several customers with very deep pockets working on their own chips to squeeze/compete with NVDA. Google, Amazon, and Microsoft all have custom ASIC and TPU projects in their pipeline, and what holds true today might not hold true in 5 years. A major reason Nvidia was able to do so well was because of technical outreach by donating their GPUs to programs all over the US, building a strong albeit self serving OSS relationship, the CUDA ecosystem, and the acquisition of Infiniband. Much of these advantages can be nullified by competitive margins and pricing for hyperscalers designed and owned hardware. Cisco was hit by this same situation as server vendors like Dell began integrating their own in-house networking functionality within their servers, and Dell itself was outcompeted by cloud vendors.
- smallmancontrov 2y ago
  Agreed, the only thing I'd add would be AMD's self-sabotage. Pretending to have compute capability (see also "right around the corner") for over a decade when it was actually so broken as to be sub-viable was an enormous own-goal that they doubled down on when they exited desperation mode without rehydrating the teams and tripled down on with the RDNA/CDNA split. By rights they should have been fast followers, but they stacked the critical mistakes so high that frankly I think NVIDIA's subsidized GPGPU classes are the smaller part of this story. If the people struggling to get OpenCL to work had been able to get it to work (on other than NVIDIA GPUs lol) the situation would look very different today.
- JKCalhoun 2y ago
  And I'd be surprised (disappointed) if AAPL were sitting on their hands.
roenxi 2y ago
I think you're not naming entire nation states, given the way the Chinese are splashing money around there it looks like there is about to be a flood of chipmaking capacity hitting the market and a lot of that will end up competing in matrix multiplying. Although I hear VCs are working on an algorithm where they burn money directly and the smoke patterns represent the solution to the multiplication. That might keep the margins high.
- smallmancontrov 2y ago
  Those Chinese companies will have to fight our western champions, TSMC in a green trenchcoat, TSMC in a red trenchcoat, TSMC in a blue trenchcoat, TSMC in a purple trenchcoat, and TSMC in a yellow trenchcoat.
  - selimthegrim 2y ago
    Is this after the same Spidermen TSMCs got down from all sitting on top of each other?
- randmeerkat 2y ago
  > Although I hear VCs are working on an algorithm where they burn money directly and the smoke patterns represent the solution to the multiplication. That might keep the margins high. This has been bothering me for some time now. It’s rather obvious to me, as an engineer, which startups are well positioned and which aren’t… Do VCs really have no standard for due diligence, do they really just not care, or is there something else at play that I’m missing…?
  - paxys 2y ago
    > It’s rather obvious to me, as an engineer, which startups are well positioned and which aren’t So I assume you are a multi-billionaire, right?
    - kasey_junk 2y ago
      You could know precisely what to invest in but not have access to the funds or deal flow and not be successful.
diamond559 2y ago
Microsoft, Meta and Oracle already spend billions on supposedly completely inferior(according to the Nvidia cultists) AMD chips. Nvidia is the biggest bubble in human history.
- 8ytecoder 2y ago
  That’s got to be Tesla. Nvidia’s PE is still half that of Tesla.
  - schmidtleonard 2y ago
    PE isn't a cheat code to rational stock valuation, it's the "nothing ever changes" assumption dressed up in a formula. PE looks at past earnings, while the price of a stock buys its future earnings. A modest PE on NVDA says "I expect datacenter revenue to continue" while a modest PE on TSLA would say "I expect robotaxi to fail." These are fair opinions to have, as are their opposites, but if datacenter revenue collapses then today's modest PE won't save NVDA and if TSLA can pull off robotaxi then today's high PE will be vindicated. Stocks are all about Future Earnings, but you can't put that in a column and sort by it so we have PE. You can certainly have success while avoiding high PE stocks, but you are on a site about startups and your name suggests you work in the tech sector, which are both places where high PE often does make sense and it pays to be familiar with the reasons.
    - marcosdumay 2y ago
      > a modest PE on TSLA would say "I expect robotaxi to fail." No. Just having competition is enough to destroy the expectations on TSLA. And realistically, they are abut last on that race, being thrown out of track about a decade ago and never managing to make anything work since then. So, yeah, betting on no competition is a very weird option.
      - JKCalhoun 2y ago
        Good point. With BMW, Lexus, Audi, Mercedes, Porsche — all with luxury EV's — Tesla really is looking like an also-ran. I'm not in that circle of clientele though so I don't know: are any of these now seen as the luxury electric car?

> And I can't name several customers with very deep pockets working on their own chips to squeeze/compete with NVDA. Google, Amazon, and Microsoft all have custom ASIC and TPU projects in their pipeline, and what holds true today might not hold true in 5 years. A major reason Nvidia was able to do so well was because of technical outreach by donating their GPUs to programs all over the US, building a strong albeit self serving OSS relationship, the CUDA ecosystem, and the acquisition of Infiniband. Much of these advantages can be nullified by competitive margins and pricing for hyperscalers designed and owned hardware. Cisco was hit by this same situation as server vendors like Dell began integrating their own in-house networking functionality within their servers, and Dell itself was outcompeted by cloud vendors.

smallmancontrov 2y ago
Agreed, the only thing I'd add would be AMD's self-sabotage. Pretending to have compute capability (see also "right around the corner") for over a decade when it was actually so broken as to be sub-viable was an enormous own-goal that they doubled down on when they exited desperation mode without rehydrating the teams and tripled down on with the RDNA/CDNA split. By rights they should have been fast followers, but they stacked the critical mistakes so high that frankly I think NVIDIA's subsidized GPGPU classes are the smaller part of this story. If the people struggling to get OpenCL to work had been able to get it to work (on other than NVIDIA GPUs lol) the situation would look very different today.
JKCalhoun 2y ago
And I'd be surprised (disappointed) if AAPL were sitting on their hands.

roenxi 2y ago

I think you're not naming entire nation states, given the way the Chinese are splashing money around there it looks like there is about to be a flood of chipmaking capacity hitting the market and a lot of that will end up competing in matrix multiplying. Although I hear VCs are working on an algorithm where they burn money directly and the smoke patterns represent the solution to the multiplication. That might keep the margins high.

smallmancontrov 2y ago
Those Chinese companies will have to fight our western champions, TSMC in a green trenchcoat, TSMC in a red trenchcoat, TSMC in a blue trenchcoat, TSMC in a purple trenchcoat, and TSMC in a yellow trenchcoat.
- selimthegrim 2y ago
  Is this after the same Spidermen TSMCs got down from all sitting on top of each other?
randmeerkat 2y ago
> Although I hear VCs are working on an algorithm where they burn money directly and the smoke patterns represent the solution to the multiplication. That might keep the margins high. This has been bothering me for some time now. It’s rather obvious to me, as an engineer, which startups are well positioned and which aren’t… Do VCs really have no standard for due diligence, do they really just not care, or is there something else at play that I’m missing…?
- paxys 2y ago
  > It’s rather obvious to me, as an engineer, which startups are well positioned and which aren’t So I assume you are a multi-billionaire, right?
  - kasey_junk 2y ago
    You could know precisely what to invest in but not have access to the funds or deal flow and not be successful.

diamond559 2y ago

Microsoft, Meta and Oracle already spend billions on supposedly completely inferior(according to the Nvidia cultists) AMD chips. Nvidia is the biggest bubble in human history.

8ytecoder 2y ago
That’s got to be Tesla. Nvidia’s PE is still half that of Tesla.
- schmidtleonard 2y ago
  PE isn't a cheat code to rational stock valuation, it's the "nothing ever changes" assumption dressed up in a formula. PE looks at past earnings, while the price of a stock buys its future earnings. A modest PE on NVDA says "I expect datacenter revenue to continue" while a modest PE on TSLA would say "I expect robotaxi to fail." These are fair opinions to have, as are their opposites, but if datacenter revenue collapses then today's modest PE won't save NVDA and if TSLA can pull off robotaxi then today's high PE will be vindicated. Stocks are all about Future Earnings, but you can't put that in a column and sort by it so we have PE. You can certainly have success while avoiding high PE stocks, but you are on a site about startups and your name suggests you work in the tech sector, which are both places where high PE often does make sense and it pays to be familiar with the reasons.
  - marcosdumay 2y ago
    > a modest PE on TSLA would say "I expect robotaxi to fail." No. Just having competition is enough to destroy the expectations on TSLA. And realistically, they are abut last on that race, being thrown out of track about a decade ago and never managing to make anything work since then. So, yeah, betting on no competition is a very weird option.
    - JKCalhoun 2y ago
      Good point. With BMW, Lexus, Audi, Mercedes, Porsche — all with luxury EV's — Tesla really is looking like an also-ran. I'm not in that circle of clientele though so I don't know: are any of these now seen as the luxury electric car?