Python 3.13 Gets a JIT • Hacker News Board

Re: Python 3.13 Gets a JIT

#551

Post by Dylan16807 » Fri, Jan 12, 2024, 6:37 AM UTC

crabbone wrote: post #410

mattgruskin wrote: post #340

Earlier quoted context omitted.

Microsoft already tried Python on the CLR! They didn't stick with it. https://en.wikipedia.org/wiki/IronPython

That's why I said another . It was a different time. Microsoft had a different strategy towards languages not developed by Microsoft. Similar to how there also used to be JScript, but now Node.js is basically a Microsoft's pet project. There are actually plenty of popular Microsoft's projects that took even more than two tries. Azure is like their third attempt at cloud services, iirc. Credit where credit is due, the…

You did not, in fact, say "another".

Re: Python 3.13 Gets a JIT

#552

Post by SpaghettiCthulu » Mon, Jan 15, 2024, 2:02 AM UTC

kzrdude wrote: post #423

SpaghettiCthulu wrote: post #313

Earlier quoted context omitted.

I believe a JIT using this technique could eliminate dead code at the Python bytecode level, but not at the machine code level. That seems pretty reasonable to me.

Not sure, these optimizations multiply in power when used together. Propagate constants and fold constants, after that you can remove things like "if 0 > 0", both the conditional check and the whole block below it, and so on.

That's entirely possible with the approach I described. Such optimizations can easily be done on the bytecode prior to the compilation to machine code.

Re: Python 3.13 Gets a JIT

#553

Post by ElectronCharge » Tue, Jan 23, 2024, 4:29 AM UTC

bee_rider wrote: post #541

ElectronCharge wrote: post #449

Earlier quoted context omitted.

The only way to achieve C/C++/Fortran efficiency is a statically compiled, strongly typed language. Witness the effort put into Java JITC and the rest of the modern Java (and Graal) runtime. Still well short of the promised “C equivalence”. To me, Mojo looks like the best approach to fusing that with the Python ecosystem! (I have no doubt about it being open sourced at some point.)

JIT and static typing are not mutually exclusive or necessarily opposed. For example, MKL added JIT for small matrix multiplications quite a while ago. https://www.intel.com/content/www/us/en/developer/articles/t... I think people just reach for JIT more often in dynamic languages because they carry more information around and have more of a performance deficit that they want to mitigate. I dunno. I mostly program in…

I meant to reply a while back, I hope you run across this.

You should definitely look into Julia. It’s a beautiful language, and squarely aimed at the Fortran space. It also relies heavily on JITC and GC. That’s fine for purely scientific computing, but not so good as a general purpose language.

That’s where C, C++, Rust, and Mojo are the current major contenders, IMO.