#2315 Community benchmark: RTX 5090 Laptop GPU + Core Ultra 9 275HX, UD-IQ4_XS, Strata 0.1.43.1

Descrição

Benchmark report from `tools/strata_bench.py` (suite 1). Proposed results folder name: `bench/results/2026-10-14-community-rtx-5090-laptop-gpu-qwen3-8-flash-next-unsloth-ud-iq4-xs`

# Community benchmark on NVIDIA GeForce RTX 5090 Laptop GPU

Measured on 2026-10-14 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 5090 Laptop GPU, 24 GiB VRAM
- CPU: Intel(R) Core(TM) Ultra 9 275HX (24 cores, 24 threads); installed RAM: 95.4 GiB; storage: not measured
- OS: Windows 10 (10.0.26300); GPU driver: not read; CUDA: not read
- Strata: project 0.1.43.1, commit not a git checkout; engine 0.1.43.1
- Background workloads and power limits: not recorded

## Model and configuration

- Model and quantization: qwen3.8-flash-next-unsloth-ud-iq4_xs
- Context: 204800 tokens; parallel slots: 1
- Settings: the install's own configuration, unchanged; thinking off for the suite; temperature 0

```text
strata.exe --pack unsloth-ud-iq4_xs --native Qwen3.8-Flash-Next-UD-IQ4_XS-00001-of-00003.gguf --expert-profile expert-profile.bin --expert-cache auto --prefill auto --spec 4 --mtp rt --max-context 204800 --kv int8 --kv-resident 32768 --resident-budget-gib 55 --vision --vram-reserve-mib 700 --pcie-frac 1.00 --spec-min-p 0.70 --pool-workers 6
```

## Method

Run by `python tools/strata_bench.py` against the HTTP API of a Strata server (already running); 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 | 51.2 (39.8-59.4) | 0.58 (0.48-0.72) |
| 4K prompt, fresh | 4019 (4019-4019) | 0 (0-0) | 24 cap | 3 | 1691.4 (915.6-1724.3) | not measured | 2.41 (2.36-4.43) |
| 16K prompt, fresh | 16058 (16058-16058) | 0 (0-0) | 24 cap | 3 | 2556.5 (2026.2-2581.7) | not measured | 6.34 (6.27-7.98) |
| agent session, follow-up turns | 14479 (14440-14528) at the last turn | see below | up to 64 | 3 | see below | 64.1 (15.6-81.7) | 2.16 (1.65-5.45) |

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: 68.8 % (68.5-68.8); earlier-turn tokens read again per session: 4 (3-144) (earlier turns that the cache did not hold; the 13635 (13635-13637) new tokens of the files are read once); first turn (nothing cached) first token: 1.47 s (1.36-1.50).

| Turn (run 1) | Prompt tokens | Reused | Read | Prompt tok/s | First token (s) | Decode tok/s |
| ---: | ---: | ---: | ---: | ---: | ---: | ---: |
| 0 | 609 | 0 | 609 | 416.8 | 1.501 | 32.1 |
| 1 | 3701 | 619 | 3082 | 570.5 | 5.453 | 48.3 |
| 2 | 5234 | 3764 | 1470 | 451.7 | 3.285 | 43.0 |
| 3 | 8678 | 5297 | 3381 | 672.2 | 5.073 | 44.4 |
| 4 | 11593 | 8733 | 2860 | 615.6 | 4.684 | 15.6 |
| 5 | 14440 | 11594 | 2846 | 736.6 | 3.91 | 55.7 |

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: 124 s.