Speed sweep
qwen38-27b-nvfp4-rtxpro6000-sglang-tp1-sweep
qwen38-27b-nvfp4-rtxpro6000-sglang-tp1-sweepRecord
- Measured
- 2026-08-24
- Accepted
- 2026-08-24
- Points
- 6
Measured speed
| Concurrency | Context | Prefill | Decode | TTFT ms | Status |
|---|---|---|---|---|---|
| 1 | 130,560 | 5,254.1 | 56.9 | 24,851.2 | accepted |
| 1 | 130,560 | 1,005,203.4 | 56.9 | 73.8 | accepted |
| 2 | 130,560 | — | 24.6 | 49,713.1 | accepted |
| 2 | 130,560 | — | 53.5 | 114.1 | accepted |
| 4 | 130,560 | — | 14.2 | 74,527 | accepted |
| 4 | 130,560 | — | 48.2 | 156.5 | accepted |
Remaining fields
Identity, launch, related records, and measured speed are shown above. This is the rest of the normalized record.
- accepted at
- 2026-08-24
- id
- qwen38-27b-nvfp4-rtxpro6000-sglang-tp1-sweep
- measured at
- 2026-08-24
- recipe id
- qwen38-27b-nvfp4-rtxpro6000-sglang-tp1
- schema version
- local-ai-registry/v1
metrics
- concurrency
- 4
- inference engine version
- 0.0.0.dev1+geec794bce
- latest point at
- 2026-08-24
- max context tokens
- 130,560
- peak prompt tps
- 1,005,203.4
- point count
- 6
rows
| concurrency | context tokens | decode tok s | decode tok s per stream | output tokens | peak vram gb | prefill tok s | samples |
|---|---|---|---|---|---|---|---|
| 1 | 130,560 | Unknown | 56.9 | 512 | Unknown | 5,254.1 | 2 |
| 1 | 130,560 | Unknown | 56.9 | 512 | Unknown | 1,005,203.4 | 2 |
| 2 | 130,560 | Unknown | 24.6 | 512 | Unknown | Unknown | 2 |
| 2 | 130,560 | Unknown | 53.5 | 512 | Unknown | Unknown | 2 |
| 4 | 130,560 | Unknown | 14.2 | 512 | Unknown | Unknown | 2 |
| 4 | 130,560 | Unknown | 48.2 | 512 | Unknown | Unknown | 2 |
Provenance & metadata (1)
source
- commit
- 9bc508601af50ddc1ff463273966fcf83bcf5b1e
- paths
- campaigns/runpod-single-gpu-2026-08-23/protocol.md, campaigns/runpod-single-gpu-2026-08-23/results.csv, campaigns/runpod-single-gpu-2026-08-23/artifacts/rtx-pro-6000-blackwell-96gb-attempt2/rtxpro6000-smart
- repository
- github.com/0xSero/inference-index ↗