NVIDIA L40 memory-bound vs compute-bound workloads

Válasz

NVIDIA L40 performance headline: 181 FP16 TFLOPS, 90.5 FP32 TFLOPS, 864 GB/s bandwidth, 48 GB VRAM.

Converted into practical benchmarks: model training a 7B-parameter LLM in FP16 with reasonable batch sizes typically saturates compute before bandwidth; real-time serving on the same model is usually bandwidth-bound and tracks the 864 GB/s figure. Diffusion image generation benchmarks sit between the two — compute-heavy steps utilise tensor cores well, while attention blocks still touch bandwidth.

Check the NVIDIA L40 page for complete specifications and related GPU matchups.