#2107 Community benchmark: RTX 6000D 84G, swift-1.5-iq2_xs, Strata 0.1.43

Description

# Community benchmark on NVIDIA RTX 6000D

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 RTX 6000D, 84 GiB VRAM
- CPU: AMD Ryzen 9 9950X 16-Core Processor (16 cores, 32 threads); installed RAM: 121.5 GiB; storage: not measured
- OS: Linux 7.0.0-31-generic; GPU driver: 595.84; CUDA: 13.2
- Strata: project 0.1.43, commit 7695859a; engine 0.1.43
- Background workloads and power limits: not recorded

## Model and configuration

- Model and quantization: swift-1.5-iq2_xs
- Context: 524288 tokens; parallel slots: 1
- Settings: the install's own configuration, unchanged; thinking off for the suite; temperature 0

```text
strata --pack swift-iq2_xs --native Swift-Qwen3.8-Flash-Next-GSQ-RCO-IQ2_XS-00001-of-00002.gguf --ple-gguf Swift-Qwen3.8-Flash-Next-GSQ-RCO-IQ2_XS-00001-of-00002.gguf --expert-profile expert-profile.bin --expert-cache auto --prefill auto --spec 4 --mtp rt --max-context 524288 --rope-scaling yarn --rope-scale 2 --kv int8 --kv-resident 32768 --vision --vram-reserve-mib 700 --spec-min-p 0.5
```

## 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 | 203.2 (179.6-272.6) | 0.15 (0.14-0.16) |
| 4K prompt, fresh | 4019 (4019-4019) | 0 (0-0) | 24 cap | 3 | 5035.7 (4573.8-5037.6) | not measured | 0.81 (0.81-0.89) |
| 16K prompt, fresh | 16058 (16058-16058) | 0 (0-0) | 24 cap | 3 | 6294.3 (6127.8-6301.7) | not measured | 2.57 (2.56-2.64) |
| agent session, follow-up turns | 14487 (14480-14491) at the last turn | see below | up to 64 | 3 | see below | 225.2 (162.5-238.9) | 0.83 (0.53-1.04) |

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 % (68.6-68.8); earlier-turn tokens read again per session: 4 (3-74) (earlier turns that the cache did not hold; the 13635 (13632-13636) new tokens of the files are read once); first turn (nothing cached) first token: 0.37 s (0.37-0.39).

| Turn (run 1) | Prompt tokens | Reused | Read | Prompt tok/s | First token (s) | Decode tok/s |
| ---: | ---: | ---: | ---: | ---: | ---: | ---: |
| 0 | 609 | 0 | 609 | 1607.3 | 0.391 | 183.5 |
| 1 | 3703 | 620 | 3083 | 3165.3 | 0.99 | 207.8 |
| 2 | 5236 | 3767 | 1469 | 2484.4 | 0.606 | 232.1 |
| 3 | 8680 | 5229 | 3451 | 3363.2 | 1.043 | 189.5 |
| 4 | 11586 | 8726 | 2860 | 3111.4 | 0.936 | 234.9 |
| 5 | 14491 | 11645 | 2846 | 3266.4 | 0.889 | 162.5 |

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: 36 s.