The Computer Architecture and Semiconductor Design Centre (CASCADE) in Cambridge/UK has 5 open PhD positions to formalize the LLVM/MLIR ecosystem. Lean and formal methods...
leanprover.zulipchat.com/#na…
#MLIR #LLVM #LeanLang #LeanProver #RISCV
Interleaved Learning and Exploration: A Self-Adaptive Fuzz Testing Framework for MLIR
#MLIR #OpenCL #Testing #Package
hgpu.org/?p=30291
📌 Development repository for the Triton language and compiler
🔹 1. **Inscription en ligne**
🔹 2. **Conférences interactives**
🔗 github.com/triton-lang/trito… #MLIR
This is a "summation" posts unpacking a unique mechanism in the Firefly compiler. With #MLIR supporting delimited continuations and interaction nets (through the DCont and Inet dialects) we can contemplate optimal machine code from high level #fsharp. speakez.tech/blog/dcont-inet…
Pretty nice sharing of experience from Google Deepmind on using #MLIR for the Shardy project: github.com/openxla/shardy
🧪 What’s next?
Modular design sets the stage for:
• Learned compiler heuristics (MLGO)
• MLIR-based heterogeneous compilation
• Pluggable backends for emerging silicon
LLVM’s backend is becoming not just reusable, but trainable.
#MLIR #MLGO #CompilerTech
Follow @DeepEngineerHub for grounded analysis of what’s happening in compiler infrastructure, language tooling, and the architectures shaping modern software.
#WebAssembly #MLIR #Compilers #PLResearch
Compilers are having a moment. MLIR is quietly becoming a common layer across otherwise unrelated toolchains.
In Deep Engineering #9:
🔹 MLIR - where it fits, and why adoption is still uneven
🔥 Mojo - new series by Ivo Balbaert
📰 Read → open.substack.com/pub/deepen…
#MLIR #MojoLang #LLVM #Compilers #DeepEngineering
2025 AsiaLLVM Developers' Meeting Talks
Videos: piped.video/playlist?list=PL…
Slides: llvm.org/devmtg/2025-06/#pro…
#LLVM #MLIR
The 2nd Innovations In Compiler Technology Workshop is here! Submit your talk proposals before 1st July 2025
Head over to compilertech.org for more details!
CC += @_hiraditya_
#compilers #llvm #mlir
🧠 MLIR (Multi-Level IR) — cutting edge compiler infrastructure designed for AI and heterogeneous hardware. Under heavy development at Google and open source communities, MLIR fuels AI model optimization and portability. #AI #Compilers #MLIR #OpenSource
Really proud of the AMD SHARK team for the MLPerf 5.0 SDxl Submission - a compute intensive workloads. A codegen'd approach with the #MLIR based #IREE compiler and the shortfin "no-gil" serving solution that is competitive.
rocm.blogs.amd.com/artificia…
Really nice in-browser notebooks introducing/exploring compiler development requiring nothing more than a web browser, using @xdslproject and #MLIR. Came about from a successful tutorial @ASPLOSConf yesterday xdsl.readthedocs.io/stable/m…
Almost 2 years at Nvidia, and the Tile IR project has been a very large part of my time here!
So happy to see it finally coming to light. The CUDA GPU driver will now include a #MLIR-based JIT compiler! :)
More MLIR-based announcement at GTC tomorrow in the Cutlass 4.0 Session!
It's great to be in sunny Monterey for #ISFPGA this week, enjoying the technical programme and looking forward to sharing our work tomorrow on seamless programming of Fortran codes on @AMDembedded #AIEngines via lowering intrinsics into the #MLIR linalg dialect.
Seamless acceleration of Fortran intrinsics via AMD AI engines
#AMD #AI #MLIR #Performance #Package
hgpu.org/?p=29780