MLGO: a Machine Learning Guided Compiler Optimizations Framework · HackerPapers discussions

MLGO: a Machine Learning Guided Compiler Optimizations Framework — mentioned by boomanaiden154

Fri, 28 Jun 2024 20:34:57 GMT

I would not say we are anywhere close to perfect in compilation. Even just looking at inlining for size, there are multiple recent studies showing ~10+% improvement (https://dl.acm.org/doi/abs/10.1145/3503222.3507744, https://arxiv.org/abs/2101.04808). There is a massive amount of headroom, and even tiny bits still matter as ~0.5% gains on code size, or especially performance, can be huge.

Paper and discussion history. A mention is not an endorsement.

MLGO: a Machine Learning Guided Compiler Optimizations Framework — mentioned by boomanaiden154

Mon, 18 Sep 2023 06:19:42 GMT

What's the benefit of having an LLM do those things in a way that guesstimates? There are big wins to be had in code-size and some wins to be had in performance related to inlining [1][2], but I think the implementation in the references directly tied into the compiler's inlining heuristic is a much better way to do that as it guarantees correctness. In addition, there's a reason that compilers basically ignore the `inline` keyword these days. For branch reordering, techniques like BOLT [5] are pretty effectively able to reorder code layout at the binary level for big performance gains by…

Paper and discussion history. A mention is not an endorsement.

MLGO: a Machine Learning Guided Compiler Optimizations Framework — mentioned by boomanaiden154

Sun, 17 Sep 2023 23:21:24 GMT

For function inlining specifically, I'm not sure LLMs are necessarily the right choice. The original MLGO paper [1] demonstrated a big code-size improvement with a ML model for making inlining decisions (7-20% code size wins), but they used tens of engineered features. Maybe a LLM could squeeze some additional size wins out, but maybe not [2]. Additionally, there are other factors to consider when productionizing these systems. Compile time is important (which LLMs will almost certainly explode), and anyone concerned about code size will probably be doing (Thin)LTO which would require feeding…

Paper and discussion history. A mention is not an endorsement.

MLGO: a Machine Learning Guided Compiler Optimizations Framework — HN submission

Tue, 01 Aug 2023 17:43:21 GMT

MLGO: A Machine Learning Guided Compiler Optimizations Framework (2021)

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

MLGO: a Machine Learning Guided Compiler Optimizations Framework — mentioned by compiler-guy

Tue, 01 Aug 2023 17:42:14 GMT

There is a fair amount of academic research on this topic. Eg, MLGO[1], but I don't know of any being used for production work. 1. https://arxiv.org/pdf/2101.04808.pdf

Paper and discussion history. A mention is not an endorsement.