io.github.hongping-zh/ecocompute-estimator · v1.0.0 · ai-and-media
EcoCompute Should-I-Quantize Estimator is healthy: it completed an MCP handshake in 534 ms and exposes 4 tools.
Energy impact of LLM quantization, calibrated to measured GPU data. Every answer states its basis.
Endpoint: https://ecocompute-estimator.zhanghongping1982.workers.dev/mcp (streamable-http)
Repository · Website · JSON · Add to my inventory
| Users (last 30 days) | 0 |
| Handshake | Healthy |
| Auth needed | None |
| Tool count | 4 |
| Latency | 534 ms |
| Last checked (UTC) | 2026-10-24T05:25:10.070Z |
| Checked (UTC) | Status | Latency | Error |
|---|---|---|---|
| 2026-10-24T05:25:10.070Z | Healthy | 534 |
estimate_quantization_energy | Estimate the task-level energy change (ΔE% vs FP16) of weight-only quantization (NF4 or INT8) for a model of a given size on a GPU architecture class, from measured-anchored curves. The answer carries basis (measured/interpolated/estimated), confidence and evidence boundaries. |
get_quantization_crossover | Get the model-size crossover for NF4/INT8 per architecture: below it quantization costs energy, above it saves. Crossovers belong to a card + software stack, not an architecture — the answer says so. |
list_data_coverage | List what is actually measured (GPU architecture × precision × size range, NVML direct), what is borrowed/estimated, and where the gaps are. Use this before trusting an estimate at the edges of coverage, and to answer 'is there data for my GPU?' honestly. |
search_measurements | Search the measured anchors by model name, GPU name, architecture or size. Returns ΔE% vs FP16 per anchor with run counts — the raw evidence behind the curves. |