Speed sweep
qwen3-32b-fp8-radeon-ai-pro-r9700-32gb-vllm-tp2-sweep
qwen3-32b-fp8-radeon-ai-pro-r9700-32gb-vllm-tp2-sweepRecord
- Measured
- 2026-05-05
- Points
- 1
Measured speed
| Concurrency | Context | Prefill | Decode | TTFT ms | Status |
|---|---|---|---|---|---|
| 4 | 32,768 | — | 19.8 | 138.3 | historical |
Remaining fields
Identity, launch, related records, and measured speed are shown above. This is the rest of the normalized record.
- id
- qwen3-32b-fp8-radeon-ai-pro-r9700-32gb-vllm-tp2-sweep
- measured at
- 2026-05-05
- recipe id
- qwen3-32b-fp8-radeon-ai-pro-r9700-32gb-vllm-tp2
- schema version
- local-ai-registry/v1
metrics
- concurrency
- 4
- inference engine version
- 0.20.1+rocm721
- latest point at
- 2026-05-05
- max context tokens
- 32,768
- peak generation tps
- 79.31
- point count
- 1
rows
| concurrency | context tokens | decode tok s | decode tok s per stream | output tokens | peak vram gb | prefill tok s | samples |
|---|---|---|---|---|---|---|---|
| 4 | 32,768 | 79.31 | 19.828 | 256 | Unknown | Unknown | 1 |
Provenance & metadata (1)
source
- repository
- www.localmaxxing.com ↗