#2046 2026-10-12-community-rtx-3060-qwen3-8-flash-next-iq3-s

Beschreibung

# Community benchmark on NVIDIA GeForce RTX 3060

Measured on 2026-10-12 with `tools/strata_bench.py` (suite 1, 3 runs per test, medians with ranges). Speeds are the engine's own timings; first-token times are measured by the client over HTTP. Speed only: no correctness check was run.

## Hardware and software

- GPU and VRAM: NVIDIA GeForce RTX 3060, 12 GiB VRAM
- CPU: Intel(R) Xeon(R) CPU E5-2680 v4 @ 2.40GHz (28 cores, 56 threads); installed RAM: 127.9 GiB; storage: NVMe3 2TB
- Motherboard: X99
- Memory: DDR4 ECC 2133Mhz
- OS: Windows 10 (10.0.19045); GPU driver: 616.92; CUDA: not read
- Strata: project 0.1.43, commit not a git checkout; engine 0.1.43
- Background workloads and power limits: not recorded

## Model and configuration

- Model and quantization: qwen3.8-flash-next-iq3_s
- Context: 131072 tokens; parallel slots: 1
- Settings: the install's own configuration, unchanged; thinking off for the suite; temperature 0

```text
strata.exe --pack iq3_s --native Qwen3.8-Flash-Next-GSQ-RCO-abliterated-IQ3_S-00001-of-00002.gguf --ple-gguf Qwen3.8-Flash-Next-GSQ-RCO-abliterated-IQ3_S-00002-of-00002.gguf --expert-profile expert-profile.bin --expert-cache auto --prefill auto:32768 --spec 4 --mtp rt --max-context 131072 --kv int8 --kv-resident 32768 --pcie-frac 0.00 --spec-min-p 0.70 --adapt-every 1 --adapt-swaps 160 --adapt-decay 0.97
```

## Method

Run by `python tools/strata_bench.py` against the HTTP API of a Strata server (started by the benchmark from the install configuration); model loading is not timed. One warm-up chat and a 1K prompt are discarded first. Every prompt starts with a unique tag, so the conversation cache cannot help the fresh-prompt tests.

- short chat: the three questions of `tools/ab_engine.py`, up to 100 new tokens, 3 runs
- 4K / 16K prompts: generated word lists sized by a calibration request, 24 new tokens, 3 fresh prompts each
- agent session: system prompt and four tools, then 5 turns that each add about 1500 words of a project file and continue from the model's own earlier answers (the way agent clients resend a conversation), 64 new tokens per turn, 3 sessions; the cache numbers come from the answer's usage (`cached_tokens`) and timings (`prompt_n`)
- concurrency (when run): N requests sent at the same moment, total tokens over the wall time of the burst, latency over all requests

## Results

| Configuration | Actual prompt tokens | Reused tokens | Generated tokens | Runs | Prompt tok/s median and range | Decode tok/s median and range | TTFT seconds median and range |
| --- | ---: | ---: | ---: | ---: | --- | --- | --- |
| short chat | 46 (43-51) | 0 | 100 (100-100) | 9 | not measured | 38.6 (34.5-48.0) | 0.89 (0.79-0.99) |
| 4K prompt, fresh | 4019 (4019-4019) | 0 (0-0) | 24 cap | 3 | 621.2 (613.7-628.2) | not measured | 6.51 (6.44-6.59) |
| 16K prompt, fresh | 16058 (16058-16058) | 0 (0-0) | 24 cap | 3 | 727.0 (724.8-731.1) | not measured | 22.14 (22.02-22.20) |
| agent session, follow-up turns | 14497 (14489-14524) at the last turn | see below | up to 64 | 3 | see below | 41.3 (34.2-51.7) | 6.05 (5.17-6.47) |

Ranges are the lowest and highest of the runs; `Runs` counts requests that returned a token.

### Agent session (conversation cache)

Cache hit over the follow-up turns: 68.9 % (68.9-68.9); earlier-turn tokens read again per session: 3 (3-4) (earlier turns that the cache did not hold; the 13634 (13632-13635) new tokens of the files are read once); first turn (nothing cached) first token: 3.14 s (3.13-3.20).

| Turn (run 1) | Prompt tokens | Reused | Read | Prompt tok/s | First token (s) | Decode tok/s |
| ---: | ---: | ---: | ---: | ---: | ---: | ---: |
| 0 | 609 | 0 | 609 | 196.7 | 3.132 | 31.5 |
| 1 | 3720 | 640 | 3080 | 509.2 | 6.094 | 40.8 |
| 2 | 5253 | 3783 | 1470 | 281.9 | 5.258 | 40.2 |
| 3 | 8697 | 5316 | 3381 | 527.0 | 6.468 | 41.3 |
| 4 | 11618 | 8760 | 2858 | 462.4 | 6.237 | 51.7 |
| 5 | 14489 | 11643 | 2846 | 469.6 | 6.116 | 41.1 |

Per-request data: `results.json`. Engine log lines from the start: `engine-log-excerpt.txt` (when this run started the server).

## Correctness and limitations

Not measured: this is a speed suite. The conversation uses the model's own replies, so token counts differ a little between runs. Machine details come from the server, `nvidia-smi`/`rocm-smi` and the OS; background load and power limits are not recorded. Total time of the suite: 249 s.