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

Nvidia Trains LLM on Chip Design

eetimes.com

Post reply on HN
Print view View original

21–30 of 151 posts

Re: Nvidia Trains LLM on Chip Design

#21

Post by drsopp »

margorczynski wrote: post #17
drsopp wrote: post #13

Earlier quoted context omitted.

I guess this would be the context window size in the case of LLMs. Edit: On second thought, maybe at a certain minimum context window size it is possible to cajole the instructions in such a way that you at any point in the process make the LLM work at a suitable level of abstraction more like humans do.

Maybe the issue is that for us the "context window" that we feed ourselves is actually a compressed and abstracted version - we do not re-feed ourselves the whole conversation but a "notion" and key points that we have stored. LLMs have static memory so I guess there is no other way as to single-pass the whole thing. For human-like learning it would need to update it state (learn) on the fly as it does inference.

Half baked idea: What if you have a tree of nodes. Each node stores a description of (a part of) a system and an LLM generated list of what the parts of it are, in terms of a small step towards concreteness. The process loops through each part in each node recursively, making a new node per part, until the LLM writes actual compilable code.

Re: Nvidia Trains LLM on Chip Design

#22

Post by falcor84 »

"""Let an ultraintelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultraintelligent machine could design even better machines; there would then unquestionably be an 'intelligence explosion,' and the intelligence of man would be left far behind... Thus the first ultraintelligent machine is the last invention that man need ever make, provided that the machine is docile enough to tell us how to keep it under control. It is curious that this point is made so seldom outside of science fiction. It is sometimes worthwhile to take science fiction seriously. """

I. J. Good, in 1965 - https://en.wikipedia.org/wiki/I._J._Good

Re: Nvidia Trains LLM on Chip Design

#23

Post by alasarmas »

RecycledEle wrote: post #18
alasarmas wrote: post #16

Earlier quoted context omitted.

Actually I think this is a good point: fundamentally an AI is forced to “color inside the lines”. It won’t tell you your business plan is stupid and walk away, which is a strong signal that is hard to ignore. So will this lead to people with more money than sense to do even more extravagantly stupid things than we’ve seen in the past, or is it basically just “Accenture-in-a-box”?

AI will absolutely rate your business plan if you ask it to. Try this prompt:"Please rate this business plan on a scale of 1-100 and provide buttle points on how it can be improved without rewriting any of it: "

I agree that AI is totally capable of rating a business plan. However, I think that the act of submitting a business plan to be rated requires some degree of humility on the part of the user, and I do doubt that an AI will “push back” when it comes to an obviously bad business plan unless specifically instructed to do so.

Re: Nvidia Trains LLM on Chip Design

#24

Post by jjoonathan »

drsopp wrote: post #21
margorczynski wrote: post #17

Earlier quoted context omitted.

Maybe the issue is that for us the "context window" that we feed ourselves is actually a compressed and abstracted version - we do not re-feed ourselves the whole conversation but a "notion" and key points that we have stored. LLMs have static memory so I guess there is no other way as to single-pass the whole thing. For human-like learning it would need to update it state (learn) on the fly as it does inference.

Half baked idea: What if you have a tree of nodes. Each node stores a description of (a part of) a system and an LLM generated list of what the parts of it are, in terms of a small step towards concreteness. The process loops through each part in each node recursively, making a new node per part, until the LLM writes actual compilable code.

Isn't that what langchain is?

Re: Nvidia Trains LLM on Chip Design

#25

Post by meltyness »

gumby wrote: post #6
hasbot wrote: post #4

The title is a bit misleading as the first sentence says "to help chip designers with tasks related to chip design, including answering general questions about chip design, summarizing bug documentation, and writing scripts for EDA tools." Still pretty cool though.

Isn’t that what chip design is?

Probably excludes qualitative tasks like architecture and apportioning resources.

Re: Nvidia Trains LLM on Chip Design

#26

Post by jdblair »

scrlk wrote: post #11
xnx wrote: post #7

Google has been using machine learning for chip design since at least 2021: https://www.nature.com/articles/s41586-021-03544-w Hasn't brought about the singularity yet.

DEC did it in the 1980s: https://en.wikipedia.org/wiki/VAX_9000#SID_Scalar_and_Vector...

SID was an "expert system with over 1000 hand-written rules"

Wow.

Re: Nvidia Trains LLM on Chip Design

#27

Post by anu7df »

taneq wrote: post #20
thechao wrote: post #10

Earlier quoted context omitted.

LLM's can't gut a fish in the cube when they get to their limits. On a more serious note: I think the high-level structuring of the architecture, and then the breakdown into tactical solutions — weaving the whole program together — is a fundamental limitation. It's akin to theorem-proving, which is just hard . Maybe it's just a scale issue; I'm bullish on AGI, so that's my preferred opinion.

> LLM's can't gut a fish in the cube when they get to their limits. Is this an idiom? Or did one of us just reach the limits of our context? :P

Office space reference.

Re: Nvidia Trains LLM on Chip Design

#29

Post by mhh__ »

alasarmas wrote: post #16
thechao wrote: post #10

Earlier quoted context omitted.

LLM's can't gut a fish in the cube when they get to their limits. On a more serious note: I think the high-level structuring of the architecture, and then the breakdown into tactical solutions — weaving the whole program together — is a fundamental limitation. It's akin to theorem-proving, which is just hard . Maybe it's just a scale issue; I'm bullish on AGI, so that's my preferred opinion.

Actually I think this is a good point: fundamentally an AI is forced to “color inside the lines”. It won’t tell you your business plan is stupid and walk away, which is a strong signal that is hard to ignore. So will this lead to people with more money than sense to do even more extravagantly stupid things than we’ve seen in the past, or is it basically just “Accenture-in-a-box”?

I wouldn't trust an absolute answer but it can help you generate counterarguments that you might miss
Post reply on HN