31–40 of 119 posts
If that's their target, then what's the point? LLVM's optimizations are not done to minimize instructions but to maximize performance. On modern processors, these can be very different things.
Chatgpt4 can do source to source optimization which is pretty cool. I got it to beat gcc at -03 on simple small toy problems. and it can do similar things with python also. But it threw its hands up when I gave it a longer piece of code to optimize.
> Our approach achieves a 3.0% improvement in reducing instruction counts over the compiler, outperforming two state-of-the-art baselines that require thousands of compilations. If that's their target, then what's the point? LLVM's optimizations are not done to minimize instructions but to maximize performance. On modern processors, these can be very different things.
But LLVM targets both depending upon what optimization pipeline you select. (-Oz/-Os are targeting minimum code size, -O1,-O2,-O3 are optimization focussed).
Code size reduction is critical in some use cases like embedded environments and mobile apps and it is a significant area of research.
Chatgpt4 can do source to source optimization which is pretty cool. I got it to beat gcc at -03 on simple small toy problems. and it can do similar things with python also. But it threw its hands up when I gave it a longer piece of code to optimize.
At my company we have some "loadbearing" code that was written by a mad scientist that no longer works at our company. We have a total ban on AI for source code analysis (so chatgpt, copilot etc are all banned). I've really wanted to throw some of the grosser parts of that codebase into gpt-4 just to see if it could bring some small amount of sanity.
Earlier quoted context omitted.
Also bugs from this approach are going to be funny - program compiled with compiler version X will work as expected and same program compiled with version X+1 will start crashing because AI under some circumstances decided that dereference of a specific pointer was unnecessary, so it won't drop it into the assembly. Good luck finding such a bug, because you will be looking on correct code, but computer will be execut…
The focus of this work is finding the optimal ordering of optimization passes to perform, not doing neural compilation. This guarantees correct code, assuming the underlying transformation passes are correct. Most work in ML for compilers focuses on replacing heuristics and phase ordering precisely because they don't impact correctness. There is some work being done on neural compilation [1], but I'm not sure that's…
lol let's say they're less likely to impact correctness than an arbitrary new optimization.
Earlier quoted context omitted.
They are not using LLM to directly produce the result code, but as tool that lists which optimisations should be done and in which order, which is fairly complex problem to solve. But if optimisation passes are implemented correctly (which is anyway required for a functioning optimising compiler), it cannot produce incorrect code, maybe only suboptimal compared to default heuristics used.
If there's a list of known optimizations that preserve correctness then it becomes an optimization problem based on output length (as a proxy for cycle count). So is the idea that an LLM is more efficient than a search or direct optimization?
Chatgpt4 can do source to source optimization which is pretty cool. I got it to beat gcc at -03 on simple small toy problems. and it can do similar things with python also. But it threw its hands up when I gave it a longer piece of code to optimize.
It does make sense that ChatGPT can beat gcc (or any compiler really); the compiler is forced to optimise code so that it still appears to work as if it's the source you wrote. Most importantly, if there are observable side-effects, compilers must preserve them, even though you know you don't actually care about them. An AI is not bound to preserve those, and can semantically change what you wrote, so long as the one…
Earlier quoted context omitted.
At my company we have some "loadbearing" code that was written by a mad scientist that no longer works at our company. We have a total ban on AI for source code analysis (so chatgpt, copilot etc are all banned). I've really wanted to throw some of the grosser parts of that codebase into gpt-4 just to see if it could bring some small amount of sanity.
Do you know why that policy is in place? Is it fear that the llm provider will steal company trade knowledge?
Earlier quoted context omitted.
It does make sense that ChatGPT can beat gcc (or any compiler really); the compiler is forced to optimise code so that it still appears to work as if it's the source you wrote. Most importantly, if there are observable side-effects, compilers must preserve them, even though you know you don't actually care about them. An AI is not bound to preserve those, and can semantically change what you wrote, so long as the one…
It seems like compilers should have an interactive mode where it suggests code that is very similar but not technically perfectly the same that is more efficient and the can accept or decline the alternate version.