Hi all!
I'm again running a BoF session on "GCC for AI" at GNU Cauldron, and I'd
like to make sure it covers what people actually want to discuss rather
than just what I've got in my head.
Some topics I'm already considering:
- ML-guided optimization passes and heuristics (e.g. inlining, register
allocation)
- Codegen and vectorization for AI/ML workloads (matmul, tensor ops)
- Target-specific codegen for AI accelerators
- Where GCC fits relative to MLIR-based toolchains for AI workloads
If you have thoughts on any of these, other topics you'd like to see
covered, or specific problems you're hitting in this space, I'd love to
hear about them, either as a reply here or off-list if you'd prefer.
I'll fold everything into the session agenda ahead of Cauldron. This
session is happening on the 3rd of October at 2pm in Room S 5 :)
Thank you,
Pietra
--
--
Pietra Ferreira (she/they)
Software Toolchain Engineer
Embecosm Limited
22 London Road,
Southampton,
England
SO15 2AF
Web:www.embecosm.com
I respect your life-work balance, there is no need to respond to this email
outside your office hours.
