#treesitter grammar for #MLIR just landed in the #LLVM repository! https://github.com/llvm/llvm-project/tree/main/mlir/utils/tree-sitter-mlir

@ril100 ECE98 an in #MLIR mit +2, Abfahrt mit +0.

OpenXLA is an open source ML compiler ecosystem which aims at bridging ML frameworks to diverse hardware: https://opensource.googleblog.com/2023/03/openxla-is-ready-to-accelerate-and-simplify-ml-development.html

#ai #ml #pytorch #tensorflow #jax #xla #mlir #foss

If there’s one thing I’ve learnt by writing a tree-sitter grammar for #MLIR, it is that MLIR is quite a mess. Makes me wonder if the project will gain traction outside its existing users.

A blog post by Google Research reflects on recent ML and computer systems work.

https://ai.googleblog.com/2023/02/google-research-2022-beyond-ml-computer.html

#tensorflow #jax #tpu #mlir

Trains spotted at Lindau-Reutin #MLIR:

- SBB RABe 503 serving as EuroCity / EuroCity-Express
- DBAG class 425 serving as InterRegio-Express
- DBAG class 612 / Bombardier RegioSwinger serving as RegionalExpress
- DBAG class 1428 / Stadler FLIRT 3 serving as RegionalExpress
- ÖBB class 4024 / Bombardier TALENT serving as Regional Express

Uuuugh, trains from three countries. That's sooo uwu 🥺👉👈. And still I knew all classes by heart 🥰.

Selbes Spiel wie beim RE96 (#MH nach #MLIR)
Im Bayrischen Teil (MH - MM) hält er nur an größeren Stationen, nach MM in BaWü hält er überall

Na es wird ja doch noch: Ankunft in #MLIR mit +4, Abfahrt mit +3, und gleich wieder ohne Maskenpflicht. #DS100 #interrail @ril100

Hm. Auch in #MH mit -1 angekommen. Jetzt noch der Pünktlichkeitstest von EC190. Steht wie immer pünktlich parat. Abfahrt 14:55 via #MLIR, #XAB, #XSSM nach #XSSG #DS100 @ril100

The program and schedule for the #fosdem 2023 #llvm devroom has been published! See https://fosdem.org/2023/schedule/track/llvm/
Presentation topics include #sycl, #compiler-rt, reducing global state, #mlir, experience of a downstream project tracking newer LLVM versions, #c2, #flang, and C/C++ embedded toolchains.

@ashafer Building a language with #MLIR should be good fun! Have a look at existing projects like Pylir and torch-mlir.

Currently reading: the Linalg dialect of #MLIR. Some of the optimization pieces are in place, but it's not yet up to a full-featured integer linear programming (ILP) solver. https://mlir.llvm.org/docs/Dialects/Linalg

Time to finally deliver my promise: I'm starting to seriously work on packaging #MLIR and #Flang for #Gentoo. For now, we're going to go with BUILD_SHARED_LIBS=ON, until upstream is ready with the final dylib API.

Giving a brief status update on the sparse compiler in MLIR during the 2022 LLVM Developers' Meeting.
#MLIR #llvm #llvmdev2022

During the MLIR summit at the 2022 LLVM Developers' Meeting, I will give a very brief progress update on the sparse compiler project here at Google (Project Updates Track). I hope to see you there!
#MLIR #llvm #sparse @Google

https://llvm.swoogo.com/2022devmtg/agenda

Here’s my #introduction.

At AMD, I lead a small team focused on high-performance 3D Graphics and #machinelearning.

I’m interested in Neural Radiance Fields, real-time super-resolution, generative models like #stablediffusion and text->3D, #gamedev, etc.

We rely on open-source software like #julialang, #pytorch, #fastai, #mlir, IREE, etc.

I live near Toronto with my wife and 2 kids.