evaluation-task · Source-linked discovery

ComputeEval: CUDA Code Generation Benchmark

Evaluates LLM capability to generate correct CUDA code for kernel implementation, memory management, and parallel algorithm optimization tasks.

Open interactive record →
OriginNVIDIATopicsGeneral capabilityStatusimported

Can support

Not independently assessed by FronteraEval yet.

Cannot support by itself

No inference beyond the upstream source should be made until the protocol is reviewed.

Original sources