dogleg77

7 ms·

1.
by dogleg77 2y ago
Nature papers get retracted every year. I have not heard of DAC papers retracted. If the Nature paper made it clear that RL is not seriously expected to work on non-TPU chips, it would have have probably been rejected. If RL works on many o
2.
by dogleg77 2y ago
The point is that the Cheng et al results and paper were shown to Google and apparently okayed by Google points of contact. After this, complaining that Cheng et al didn't ask someone outside Google makes little sense. These far fetche
3.
by dogleg77 2y ago
Chip designers run EDAtools on premises. How would EDA companies have access to customer data? Maybe for debugging purposes under NDA, but that won't allow training on customer IP.
4.
by dogleg77 2y ago
No, it doesn't suggest that. Complaints are often dismissed on technicalities or because they are written poorly. Google claimed their new algorithm as a breakthrough. If this were the so, the algorithm would have helped design chips i
5.
by dogleg77 2y ago
That is misleading. The first two authors left Google in August 2022 under unclear circumstances. The code and data were owned by Google, that's probably why Kahng continued discussibg code and data with his Google contacts. He receive
6.
by dogleg77 2y ago
I am referring to direct comparisons in Cheng et al and in Stronger Baselines that everyone is discussing. Let's assume your point about "might have been insufficient". We don't currently have the luxury to be frequentis
7.
by dogleg77 2y ago
At least they are acknowledging skepticism now. That's a step in the right direction.
8.
by dogleg77 2y ago
Markov's article has been on arxiv since 2023 before the perceived conflict. https://arxiv.org/abs/2306.09633 All his affiliations have been disclosed and likely known to Jeff Dean. Jeff is simply unable to res
9.
by dogleg77 2y ago
You must be kidding. They did reach out and documented their interactions, with names of engineers, dates, and all. Someone is ghosting someone here.
10.
by dogleg77 2y ago
How would I know either way? Someone mentioned here before that Google folks have been using "hyperbolae". So, if MediaTek can clarify how they are using AlphaChip, everyone wins.
11.
by dogleg77 2y ago
Please specify what you mean by "had a hand in TPU" and where else he had his hand. Thank you
12.
by dogleg77 2y ago
Something is off with your source references. Chatterjee's amended complaint is a legal document filed under penalty of perjury, accepted by a US judge, available publicly, apparently not "thrown out". How does an earlier doc
13.
by dogleg77 2y ago
The Google internal paper by Chatterjee and the Cheng et al paper from UCSD made such comparisons with Simulated Annealing. The annealer in the Nature paper was easy to improve. When given the same time budget, the improved annealer produce
14.
by dogleg77 2y ago
I am trying to understand what you mean here by potential to disrupt. AlphaChip addresses one out of hundreds of tasks in chip design. Macro placement is a part of mixed-size placement, which is handled just fine by existing tools, many aca
15.
by dogleg77 2y ago
What did the TPU physical design team say about this publicly? Can you also point to a statement from MediaTek? (I've seen a quote in Google blog, but was unable to confirm it). Who in the TF-agents team has serious physical design bac
16.
by dogleg77 2y ago
Reading those papers and looking at the code, it doesn't look easy. However, let's imagine that the Cheng et al team comes back with results for pretraining a few months from now, and they support the conclusions of their earlier
17.
by dogleg77 2y ago
When Google published the Nature article, Nature included a rosy intro article by a leading expert in chip design. His name was Andrew Kahng, and he apparently liked Google at the time. But when he dug into Google code (released way after p
18.
by dogleg77 2y ago
Good point, but Google is buying Nvidia GPUs for some reason. Please remind me who's buying TPUs.
19.
by dogleg77 2y ago
That made sense when Jeff Dean gave talks in 2020, 2021, and 2022. He is now responding to skepticism from the EDA community by unscholarly personal attacks and vague references to "many companies" using the work. He is beyond ben
20.
by dogleg77 2y ago
That's an appeal to authority, and not an effective one. Jeff Dean doesn't have a good track record in chip design.
21.
by dogleg77 2y ago
Yes, that's unfortunate. In that case, the Google Nature paper doesn't belong in Nature. Google is definitely free to claim that they revolutionized their business, but without scientific evidence this is just marketing.
22.
by dogleg77 2y ago
Fair point. I looked at the addendum, and it doesn't really address the critiques. They show an "ablation study" on one additional circuit without describing how big it is, etc. The few numbers they give suggest that this cir
23.
by dogleg77 2y ago
Science is not done by quotes from VPs, and we don't know how MediaTek used these methods. Also, would you like to hear from VPs who wasted their company resources on Google RL and gave up? The more marketing claims we see, the less co
24.
by dogleg77 2y ago
The problem with the Google Nature paper is that its results were not reproduced outside Google. You can attack attempts to reproduce but that only reinforces the point: those claimed results cannot be trusted. Other commenters already addr