hard20 pts · first answer countsmodel-servingllm-inferenceperformance
Estimating prefill time
A 7B model processes a 2,000-token prompt. Forward compute is about 2 FLOPs per parameter per token. If the GPU sustains 400 TFLOPS, roughly how long does prefill take (ignore attention)?
Discussion 0
Join free to post in the discussion.
Related questions
- When batch inference fitsMLOps Engineer
- Traffic before the model has loadedMLOps Engineer
- Why export a model to ONNXMLOps Engineer
- Protocol for internal inference callsMLOps Engineer
- Latency and throughputMLOps Engineer
- What TTFT measuresMLOps Engineer
- Gap between streamed tokensMLOps Engineer
- Two phases of LLM generationMLOps Engineer