Rocm

AMD's MI450 Gluon-based attention decode kernel achieves up to ~85% of peak HBM bandwidth (about 16–17 TB/s of a 20 TB/s peak) using a four-stage pipeline, TDM data paths, and Split-k across KV partitions, backed by detailed hardware features and open-source references.

AMD unveils a high-end data-center stack (Helios 72-GPU rack with MI455) and a broad CPU/GPU roadmap, backed by major Anthropic/OpenAI deals and a Cerebras disaggregate-inference play, signaling a meaningful shift toward AMD-enabled AI infra dominance by 2027-2028.

CUDA creates Nvidia’s software moat, making it effectively a software company with ecosystem lock-in that heightens competitive barriers and pricing power.