2026-10-11 16:36 UTC

inference-kernels

band: coolmomentum: stable score: 0.032
temperature history

Episodes (2)

Independent benchmarks will determine whether the reported open kernels can sustain roughly 78,500 output tokens per second for Qwen3.6-35B-A3B on eight AMD MI350X GPUs under practically comparable serving conditions.
expiredknownscott: low
MoA's author claims its published attention-kernel project establishes minimality before implementation, potentially providing a proof-first basis for kernel design rather than relying solely on empirical optimization.
seedknownscott: low