Hot Chips 2026: CUDA Targets RISC-V – By Chester Lam
rbanffy
85 points
10 comments
August 24, 2026
Related Discussions
Found 5 related stories in 54.8ms across 4,281 title embeddings via pgvector HNSW
- Book: RISC-V System-on-Chip Design xlmnxp · 12 pts · July 09, 2026 · 59% similar
- RISC-V: They should have known better kaycebasques · 155 pts · August 14, 2026 · 57% similar
- Hot Chips 2026: Samsung and HBM Base Die Opportunities rbanffy · 22 pts · August 24, 2026 · 57% similar
- RISC-V Is Inevitable: State of the Union Keynote Argues signa11 · 110 pts · July 15, 2026 · 57% similar
- AMD publishes machine-readable ISA so frontier models can write its GPU kernels logickkk1 · 16 pts · July 25, 2026 · 56% similar
Discussion Highlights (6 comments)
nozzlegear
All these AI do is eat hot chip and lie
camel-cdr
related: https://www.sifive.com/development-platforms/sifive-bigsky-s...
LogTrim
The interesting part is that this is less "CUDA now runs on RISC-V" and more "Nvidia is defining what a CUDA-capable RISC-V server has to look like." That could end up being a pretty influential de facto server profile for the architecture.
djmips
hot chips. hot planet.
varispeed
Just make more RAM.
einpoklum
CUDA is proprietary technology, guarded by patents and mostly closed-source. And its originator, NVIDIA, while being a member of the Khronos consortium, and technically supporting OpenCL, makes it much more difficult for you to do anything in OpenCL than in CUDA - even though it is sometimes almosy zero effort to enable some OpenCL equivalent (e.g. : Half-precision types. They exist in CUDA; in OpenCL on NVIDIA - you can't have them). So, live with CUDA on other hardware? No thank you.