Azalia Mirhoseini, Anna Goldie, Mustafa Yazgan
Joe Wenjie Jiang, Ebrahim Songhori, Shen Wang, Young-Joon Lee, Eric Johnson, Omkar Pathak, Azade Nova, Jiwoo Pak, Andy Tong, Kavya Srinivasa, William Hang, Emre Tuncer, Quoc V. Le, James Laudon, Richard Ho, Roger Carpenter, Jeff Dean
2021 · Nature
Original HN comments, with labels from simple text rules. A label describes the whole comment and may miss context or irony. Read the discussion to make up your own mind. About the labels
6 selected excerpts · 6 total mentions
The original paper reports P&R metrics (WNS, TNS, area, power, wirelength, horizontal congestion, vertical congestion) - https://www.nature.com/articles/s41586-021-03544-w (no paywall): https://www.cl.cam.ac.uk/~ey204/teaching/ACS/R244_2021_2022/...
This work from Google (original Nature paper: https://www.nature.com/articles/s41586-021-03544-w) has been credibly criticized by several researchers in the EDA CAD discipline. These papers are of interest: - A rebuttal by a researcher within Google who wrote this at the same time as the "AlphaChip" work was going on ("Stronger Baselines for Evaluating Deep Reinforcement Learning in Chip Placement"): http://47.190.89.225/pub/education/MLcontra.pdf - The 2023 ISPD paper from a group at UCSD ("Assessment of Reinforcement Learning for Macro Placement"): https://vlsicad.ucsd.edu/Publications/Conferences/396/c396.p... - A paper from Igor Markov which critically evaluates the…
"AlphaChip" algorithm ("The False Dawn: Reevaluating Google's Reinforcement Learning for Chip Macro Placement"): https://arxiv.org/pdf/2306.09633 In short, the Google authors did not fairly evaluate their RL macro placement algorithm against other SOTA algorithms: rather they claim to perform better than a human at macro placement, which is far short of what mixed-placement algorithms are capable of today. The RL technique also requires significantly more compute than other algorithms and ultimately is learning a surrogate function for placement iteration rather than learning any novel representation of the placement problem itself. In full disclosure, I am quite skeptical of their work and wrote a detailed post on my website: https://vighneshiyer.com/misc/ml-for-placement/
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.
Past week there was news of: low power risc-v gpus, fpga-integrated sdr, aws quantum internet. It feels like the alien tech is slipping out the back door of the lab ;) Google AI beats humans at designing computer chips https://www.nature.com/articles/s41586-021-03544-w
I remember reading the original paper [0] when it came out last year, and honestly I found it pretty weak compared to other google or deepmind papers, particularly the comparison to Simulated Annealing methods and the lack of data to enable reproducibility. Felt more like a marketing paper based on proprietary data (the paper reported how they were able to produce an efficient floorplan for one of google's internal processors). Point being that aside from the controversy being stoked based on previous firings, I'm personally curious about the results showing problems on the original paper (whose senior author was Jeff Dean btw) [0] https://www.nature.com/articles/s41586-021-03544-w (the endgadget story has a link to the full pdf)
An AlexNet/ResNet-type moment may be in the cards for GAs, but I wouldn't put any money on it. They're typically only one better than brute-force. This can be good enough (and is certainly easy to implement), but if you can get a gradient for your problem -- you should use that. And nowadays, you can typically get a gradient! Most recent advances in the fields you mentioned were driven by gradient-based optimization (e.g., drug design, routing, or chip design: https://www.nature.com/articles/s41586-021-03544-w). Nature can't SGD through genomes but has a metric ton of time, so evolution might be near-ideal for sexual reproduction. We typically don't have billions of…
generations, trillions of instantiations, and complex environments to play with when optimizing functions... It's telling that the fastest-evolving biological system (our brain!) certainly doesn't employ large-scale GA; if anything, it probably approximates gradients via funky distributed rules. EDIT: The most modern application I can think of was some stuff from OpenAI (https://openai.com/blog/evolution-strategies/). But the point here is one of computational feasibility -- if they could backprop through the same workload, they would.
More from the same topics.
Dan Tong, David J. Farnham, Lei Duan +4
Abstract If future net-zero emissions energy systems rely heavily on solar and wind resources, spatial and temporal mismatches between resource availability and electricity demand may challenge system reliability.
Steven M. Paul, Daniel S. Mytelka, Christopher T. Dunwiddie +4