LLM Prefill
llm_prefill is a throughput-oriented diagnostic workload for long prompts. It
combines projection-like math with tiled causal scans over a synthetic context
store, creating a different memory and compute mix from token-by-token decode.
| Area | What the workload exercises |
|---|---|
| Context store | A synthetic context store sized from free VRAM. |
| Attention path | Tiled causal scans through the context window. |
| Compute | Projection and attention-score-like FMA work. |
| Result | Synthetic prompt token iterations per second, reported as prompt-tokens/s. |
The workload helps isolate how longer contexts affect cache, DRAM, clocks, power, and thermals. It is not a substitute for measuring a particular model, attention implementation, precision, batching policy, or serving engine.