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

Nvidia's $20B antitrust loophole

ossa-ma.github.io

Post reply on HN
Print view View original

21–30 of 184 posts

Re: Nvidia's $20B antitrust loophole

#21

Post by ossa-ma »

vineethy wrote: post #14

I think it's important to note that there's nothing forbidding LPU style determinism from being used in training. They just didn't make that choice. Also tenstorrent could be a viable challenger in this space. It seems to me that their NoC and their chips could be mostly deterministic as long as you don't start adding in branches

You're right but my understanding is that Groq's LPU architecture makes it inference-only in practice.

Like Groq's chips only have 230MB of SRAM per chip vs 80GB on an H100, training is memory hungry as you need to hold model weights + gradients + optimizer states + intermediate activations.

Re: Nvidia's $20B antitrust loophole

#23

Post by bionhoward »

vineethy wrote: post #14

I think it's important to note that there's nothing forbidding LPU style determinism from being used in training. They just didn't make that choice. Also tenstorrent could be a viable challenger in this space. It seems to me that their NoC and their chips could be mostly deterministic as long as you don't start adding in branches

Would SRAM make weight updates prohibitive vs DRAM?

Re: Nvidia's $20B antitrust loophole

#24

Post by yesco »

yalogin wrote: post #11

Nvidia bought groq? I am out of the loop but what does groq have that they want?

I'll just make note here for anyone else confused that Groq and Grok are distinct entities. They just have similar names.

Groq is more of a hardware focused company.

Re: Nvidia's $20B antitrust loophole

#26

Post by some_guy_nobel »

georgeburdell wrote: post #25

This behavior is extremely damaging to the startup scene. Who would join a startup these days unless it’s run by a close friend or relative? At least in that case, the scorned junior employees would have social recourse.

Right? As a former founder, I laugh every time I get a 'Founding Engineer' recruit email...

Re: Nvidia's $20B antitrust loophole

#27

Post by websiteapi »

I wonder how the startup scene will adjust to this if it becomes mainstream. can employee contracts be modified to force compensation even in this case? seems difficult to write one up without weird second order effects.

if this does end up being something that is legal and successfully circumvents anti trust, does it mean antitrust actually is a failure in practice?

2026 hasn't even begun and more shenanigans are in flight.

Re: Nvidia's $20B antitrust loophole

#28

Post by paxys »

georgeburdell wrote: post #25

This behavior is extremely damaging to the startup scene. Who would join a startup these days unless it’s run by a close friend or relative? At least in that case, the scorned junior employees would have social recourse.

Yeah unless you are a founder or top investor it’s pretty much a guarantee that there will be no exit.

Re: Nvidia's $20B antitrust loophole

#29

Post by evilduck »

georgeburdell wrote: post #25

This behavior is extremely damaging to the startup scene. Who would join a startup these days unless it’s run by a close friend or relative? At least in that case, the scorned junior employees would have social recourse.

I don't think this makes it much worse because that's hard to do, it's already terrible. Getting screwed by startup founders has been the status quo for at least 15 or 20 years now.

If you're just a worker then demand fair market wages, work healthy hours, and treat your useless class of shares as already used and discarded scratch off lottery tickets.

Post reply on HN