4 ms·

Read up a bit on the effort needed to get a fab going, and the yield rates. While engineers are crucial in the setup, the fab itself is not as 'fungible' as the

by NBJack 1y ago

Read up a bit on the effort needed to get a fab going, and the yield rates. While engineers are crucial in the setup, the fab itself is not as 'fungible' as the employees involved. I can spin up a strong ML team through hiring in probably 6-12 months with the right funding. Building a chip fab and getting it to a sensible yield would take 3-5 years, significantly more funding, strong supply lines, etc.

singron 1y ago
Nvidia isn't a fab.
wongarsu 1y ago
But the fabs don't belong to NVIDIA, they belong to TSMC. I have no doubt that Taiwan and maybe even the US government would step in to save TSMC if for some reason it got existential problems, but that doesn't provide an argument for saving NVIDIA
trollbridge 1y ago
Right. I could spin up a strong ML team, an AI startup, build a foundational model, etc give a reasonable amount of seed capital. Build a chip fab? I’ve got no idea where to start, where to even find people to hire, and i know the equipment we’d need to acquire would be also quite difficult to get at any price.
OfficialTurkey 1y ago
> I can spin up a strong ML team through hiring in probably 6-12 months with the right funding. Mark Zuckerberg would like a word with you
embedding-shape 1y ago
> I can spin up a strong ML team through hiring in probably 6-12 months with the right funding Not sure what to call this except "HN hubris" or something. There are hundreds of companies who thought (and still think) the exact same thing, and even after 24 months or more of "the right funding" they still haven't delivered the results. I think you're misunderstanding how difficult all of this is, if you think it's merely a money problem. Otherwise we'd see SOTA models from new groups every month, which we obviously aren't, we have a few big labs iteratively progressing SOTA, with some upstarts appearing sometimes (DeepSeek, Kimi et al) but it isn't as easy as you're trying to make it out to be.
- whimsicalism 1y ago
  There’s a lot in LLM training that is pretty commodity at this point. The difficulty is in data - and a large part of why it has gotten more challenging is simply that some of the best sources of data have locked down against scraping post-2022 and it is less permissible to use copyrighted data than the “move fast and break things” pre-2023 era. As you mentioned, multiple no name chinese companies have done it and published many of their results. There is a commodity recipe for dense transformer training. The difference between Chinese and US is that they have less data restrictions. I think people overindex on the Meta example. It’s hard to fully understand why Meta/llama have failed as hard as they have - but they are an outlier case. Microsoft AI only just started their efforts in earnest and are already beating Meta shockingly.
- marcyb5st 1y ago
  Fully agree. I also think we are deep into the diminishing returns territory. If I have to guess OAI and others pay top dollars for talent that has a higher probability of discovering the next "attention" mechanism and investors are betting this is coming soon (hence the hige capitalizations and willing to loive with 11B losses/quarter). If they lose patience in throwing money at the problem I see only few players remaining in the race because they have other revenue streams
- noosphr 1y ago
  >Otherwise we'd see SOTA models from new groups every month We do. It's just that startups don't go after the frontier models but niche spaces which are under served and can be explored with a few million in hardware. Just like how open AI made gpt2 before they made gpt3.
  - embedding-shape 1y ago
    > We do. > It's just that startups don't go after the frontier models but niche spaces But both of "New SOTA models every month" and "Startups don't go for SOTA" cannot be true at the same time. Either we get new SOTA models from new groups every month (not true today at least) or we don't, maybe because the labs are focusing on non-SOTA instead.

by NBJack 1y ago

singron 1y ago

Nvidia isn't a fab.

wongarsu 1y ago

But the fabs don't belong to NVIDIA, they belong to TSMC. I have no doubt that Taiwan and maybe even the US government would step in to save TSMC if for some reason it got existential problems, but that doesn't provide an argument for saving NVIDIA

trollbridge 1y ago

Right. I could spin up a strong ML team, an AI startup, build a foundational model, etc give a reasonable amount of seed capital. Build a chip fab? I’ve got no idea where to start, where to even find people to hire, and i know the equipment we’d need to acquire would be also quite difficult to get at any price.

OfficialTurkey 1y ago

> I can spin up a strong ML team through hiring in probably 6-12 months with the right funding. Mark Zuckerberg would like a word with you

embedding-shape 1y ago

> I can spin up a strong ML team through hiring in probably 6-12 months with the right funding Not sure what to call this except "HN hubris" or something. There are hundreds of companies who thought (and still think) the exact same thing, and even after 24 months or more of "the right funding" they still haven't delivered the results. I think you're misunderstanding how difficult all of this is, if you think it's merely a money problem. Otherwise we'd see SOTA models from new groups every month, which we obviously aren't, we have a few big labs iteratively progressing SOTA, with some upstarts appearing sometimes (DeepSeek, Kimi et al) but it isn't as easy as you're trying to make it out to be.

whimsicalism 1y ago
There’s a lot in LLM training that is pretty commodity at this point. The difficulty is in data - and a large part of why it has gotten more challenging is simply that some of the best sources of data have locked down against scraping post-2022 and it is less permissible to use copyrighted data than the “move fast and break things” pre-2023 era. As you mentioned, multiple no name chinese companies have done it and published many of their results. There is a commodity recipe for dense transformer training. The difference between Chinese and US is that they have less data restrictions. I think people overindex on the Meta example. It’s hard to fully understand why Meta/llama have failed as hard as they have - but they are an outlier case. Microsoft AI only just started their efforts in earnest and are already beating Meta shockingly.
marcyb5st 1y ago
Fully agree. I also think we are deep into the diminishing returns territory. If I have to guess OAI and others pay top dollars for talent that has a higher probability of discovering the next "attention" mechanism and investors are betting this is coming soon (hence the hige capitalizations and willing to loive with 11B losses/quarter). If they lose patience in throwing money at the problem I see only few players remaining in the race because they have other revenue streams
noosphr 1y ago
>Otherwise we'd see SOTA models from new groups every month We do. It's just that startups don't go after the frontier models but niche spaces which are under served and can be explored with a few million in hardware. Just like how open AI made gpt2 before they made gpt3.
- embedding-shape 1y ago
  > We do. > It's just that startups don't go after the frontier models but niche spaces But both of "New SOTA models every month" and "Startups don't go for SOTA" cannot be true at the same time. Either we get new SOTA models from new groups every month (not true today at least) or we don't, maybe because the labs are focusing on non-SOTA instead.