- First word afterttft_median_s · s
- 467.26 sper person · median of 5
- Slowest first wordttft_max_s · s
- 467.26 sper person · max of 5
- Machine totalaggregate_decode_tps · tok/s
- 36.53 tok/smachine total · single · sample count not recorded
- Machine total including first-word waitaggregate_e2e_tps · tok/s
- 1.13 tok/smachine total · single · sample count not recorded
- Everyone answered inwall_seconds · s
- 482.07 smachine total · single · sample count not recorded
- Peak memory, whole hostpeak_host_unavailable_gib · GiB
- 103.59 GiBmachine total · max · sample count not recorded
Runs on this setup
Earlier runs keep their own engine versions and outcomes. Workload shapes stay separate.
History and measured sweep
| Date | Shape | Each person · median | Machine total | First word · median | Result | Run |
|---|---|---|---|---|---|---|
2026-10-06 | 5 people · shared · warm | not recorded | 24.22 tok/s | 10.88 s | Passed | |
2026-10-06 | 5 people · shared · cold | not recorded | 35.43 tok/s | 467.14 s | Passed | |
2026-10-06 | 5 people · shared · warm | not recorded | 25.58 tok/s | 8.91 s | Passed | |
2026-10-06 | 5 people · shared · cold | not recorded | 36.53 tok/s | 467.26 s | Passed | |
2026-10-06 | 5 people · distinct · cold | not recorded | 1.67 tok/s | 243.79 s | Passed | |
2026-10-06 | 5 people · distinct · cold | not recorded | 0.72 tok/s | 552.72 s | Passed | |
2026-10-06 | 5 people · distinct · cold | not recorded | 0.28 tok/s | 1,395.27 s | Failed | |
2026-10-06 | 4 people · distinct · cold | not recorded | 0.29 tok/s | 1,162.91 s | Passed | |
2026-10-06 | 3 people · distinct · cold | not recorded | 0.33 tok/s | 932.03 s | Passed | |
2026-10-06 | 2 people · distinct · cold | not recorded | 0.44 tok/s | 695.60 s | Passed | |
2026-10-06 | 1 people · distinct · cold | not recorded | 9.05 tok/s | 457.98 s | Passed | |
2026-10-06 | 2 people · distinct · cold | not recorded | not recorded | not recorded | Failed | |
2026-10-06 | 1 people · distinct · cold | not recorded | not recorded | not recorded | Failed |
Each person in this run
First-word wait against writing speed after that word, including pauses. Earlier streams can pause while other people’s prompts are read. The table preserves filled input, cached-token counts and each person’s checks.
Per-person table · all measured request rows
| Person | Filled input | Cached tokens | Output | First word | Writing after first word | Total time | Checks | Run |
|---|---|---|---|---|---|---|---|---|
3 | 250,011 | unknown | 109 | 467.26 s | 7.31 tok/sone person · includes waits | 482.03 s | 3/3 Hidden facts found · completed · checks passed | |
2 | 250,011 | unknown | 109 | 467.26 s | 7.31 tok/sone person · includes waits | 482.03 s | 3/3 Hidden facts found · completed · checks passed | |
4 | 250,011 | unknown | 109 | 467.26 s | 7.31 tok/sone person · includes waits | 482.03 s | 3/3 Hidden facts found · completed · checks passed | |
1 | 250,011 | unknown | 109 | 467.26 s | 7.31 tok/sone person · includes waits | 482.04 s | 3/3 Hidden facts found · completed · checks passed | |
5 | 250,011 | unknown | 109 | 467.26 s | 7.31 tok/sone person · includes waits | 482.04 s | 3/3 Hidden facts found · completed · checks passed |
Per-person table · all measured request rows
| Person | Filled input | Cached tokens | Output | First word | Writing after first word | Total time | Checks | Run |
|---|---|---|---|---|---|---|---|---|
3 | 250,011 | unknown | 109 | 467.26 s | 7.31 tok/sone person · includes waits | 482.03 s | 3/3 Hidden facts found · completed · checks passed | |
2 | 250,011 | unknown | 109 | 467.26 s | 7.31 tok/sone person · includes waits | 482.03 s | 3/3 Hidden facts found · completed · checks passed | |
4 | 250,011 | unknown | 109 | 467.26 s | 7.31 tok/sone person · includes waits | 482.03 s | 3/3 Hidden facts found · completed · checks passed | |
1 | 250,011 | unknown | 109 | 467.26 s | 7.31 tok/sone person · includes waits | 482.04 s | 3/3 Hidden facts found · completed · checks passed | |
5 | 250,011 | unknown | 109 | 467.26 s | 7.31 tok/sone person · includes waits | 482.04 s | 3/3 Hidden facts found · completed · checks passed |
Answer checks
15/15 Hidden facts found5/5 request checks
Method
Synchronized long public-domain prompts, three exact hidden facts per user. Decode rate uses completion tokens minus one over first output through stream completion; effective prefill includes scheduling and first-token latency. Aggregate decode uses the union window from first output to final completion; end-to-end rate includes prefill.
Limits
Lab observation, not customer acceptance. Prompt content length excludes the server chat template; actual per-request usage is recorded. Synthesis is requested and retained but semantic accuracy is unassessed. All requested users must overlap in observed decode to pass.
What ran
Exact identity
- Original publisher model
- Qwen3.8-27B
- Publisher revision
1d4bf0f2ff6012fd82039f2fa52739d0dd7c60c0- Variant · format · precision
- INT4 variant · safetensors · INT4
- Weights SHA-256
1d8268aa85ace093a561e3e7b63b9d390dac1cd55a90cd55b5ec509c3c9da9fef078f51aed0c1dba613c1759487004ee06aba1a290606b6be125f59ab710959e- Engine · version · feature
- NVIDIA NGC vLLM ·
0.29.0+5013de39.dev· plain - Hardware
- NVIDIA GB10, 128 GB unified memory
- Harness / workload
keep-long-context-concurrency-v1 · 1- Workload suite · version
not recorded- Harness SHA-256
not recorded- Fixture SHA-256
2d23afb10cda3f1d9b23f9375d8c3456fc82a4da61760dec43722fa7b44f92f8
Workload shape
- cache state
- cold
- People in this run
- 5
- Configured window (tokens)
- 262144
- Filled content target (tokens)
- 250000
- fixture variant
- shared-question-order-v1
- KV cache precision
- fp8
- Output token ceiling
- 512
- scenario
- shared
- Thinking
- off
Full reproduction pack available to customers and evaluators on request
All recorded numbers
Metric definitions, scope and statistics
| Metric | Value | Scope | Statistic | Sample count | Basis / source field | Run |
|---|---|---|---|---|---|---|
wall seconds | 482.07 s | machine total | single | not recorded | wall_seconds | |
ttft median s | 467.26 s | per person | median | 5 | ttft_median_s | |
hidden facts total | 15.00 count | machine total | single | not recorded | needle_total | |
ttft max s | 467.26 s | per person | max | 5 | ttft_max_s | |
requests ok | 5.00 count | machine total | single | not recorded | requests_ok | |
hidden facts correct | 15.00 count | machine total | single | not recorded | needle_correct | |
decode overlap | 5.00 count | machine total | single | not recorded | decode_overlap | |
aggregate concurrency | 5.00 count | machine total | single | not recorded | aggregate_concurrency | |
requests total | 5.00 count | machine total | single | not recorded | requests_total | |
aggregate decode tps | 36.53 tok/s | machine total | single | not recorded | aggregate_tps | |
aggregate e2e tps | 1.13 tok/s | machine total | single | not recorded | aggregate_end_to_end_tps | |
peak host unavailable gib | 103.59 GiB | machine total | max | not recorded | peak_host_unavailable_gib |
Data and citation
Skilak Keep evidence, run 28597e232406401b9487303b, measured 2026-10-06T04:49:28.129613+00:00. https://skilak.ai/keep/resources/evidence/r/28597e232406401b9487303b
