- First word afterttft_median_s · s
- 140.65 sper person · median of 5
- Slowest first wordttft_max_s · s
- 239.41 sper person · max of 5
- Machine totalaggregate_decode_tps · tok/s
- 2.62 tok/smachine total · single · sample count not recorded
- Machine total including first-word waitaggregate_e2e_tps · tok/s
- 2.13 tok/smachine total · single · sample count not recorded
- Everyone answered inwall_seconds · s
- 248.80 smachine total · single · sample count not recorded
- Peak memory, whole hostpeak_host_unavailable_gib · GiB
- 114.20 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 | 33.59 tok/s | 10.03 s | Passed | |
2026-10-06 | 5 people · shared · cold | not recorded | 19.45 tok/s | 373.65 s | Failed | |
2026-10-06 | 5 people · distinct · cold | not recorded | 2.62 tok/s | 140.65 s | Failed | |
2026-10-06 | 5 people · distinct · cold | not recorded | 1.06 tok/s | 372.46 s | Failed | |
2026-10-06 | 5 people · distinct · warm | not recorded | 0.47 tok/s | 726.13 s | Failed | |
2026-10-06 | 5 people · distinct · cold | not recorded | 0.37 tok/s | 1,082.95 s | Failed | |
2026-10-06 | 4 people · distinct · cold | not recorded | 0.36 tok/s | 914.14 s | Passed | |
2026-10-06 | 3 people · distinct · cold | not recorded | 0.41 tok/s | 732.45 s | Passed | |
2026-10-06 | 2 people · distinct · cold | not recorded | 0.53 tok/s | 542.39 s | Passed | |
2026-10-06 | 1 people · distinct · cold | not recorded | 8.68 tok/s | 360.53 s | Passed |
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 | 64,011 | unknown | 102 | 48.35 s | 0.69 tok/sone person · includes waits | 193.91 s | 3/3 Hidden facts found · completed · checks passed | |
4 | 64,011 | unknown | 107 | 94.46 s | 0.73 tok/sone person · includes waits | 239.78 s | 3/3 Hidden facts found · completed · checks passed | |
2 | 64,011 | unknown | 108 | 181.70 s | 1.84 tok/sone person · includes waits | 239.90 s | 3/3 Hidden facts found · completed · checks passed | |
1 | 64,011 | unknown | 110 | 140.65 s | 1.10 tok/sone person · includes waits | 240.11 s | 3/3 Hidden facts found · completed · checks passed | |
5 | 64,011 | unknown | 104 | 239.41 s | 10.99 tok/sone person · includes waits | 248.79 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 | 64,011 | unknown | 102 | 48.35 s | 0.69 tok/sone person · includes waits | 193.91 s | 3/3 Hidden facts found · completed · checks passed | |
4 | 64,011 | unknown | 107 | 94.46 s | 0.73 tok/sone person · includes waits | 239.78 s | 3/3 Hidden facts found · completed · checks passed | |
2 | 64,011 | unknown | 108 | 181.70 s | 1.84 tok/sone person · includes waits | 239.90 s | 3/3 Hidden facts found · completed · checks passed | |
1 | 64,011 | unknown | 110 | 140.65 s | 1.10 tok/sone person · includes waits | 240.11 s | 3/3 Hidden facts found · completed · checks passed | |
5 | 64,011 | unknown | 104 | 239.41 s | 10.99 tok/sone person · includes waits | 248.79 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 did not work
- Observed decode overlap was 4 of 5 requested users.
What ran
Exact identity
- Original publisher model
- Qwen3.8-27B
- Publisher revision
1d4bf0f2ff6012fd82039f2fa52739d0dd7c60c0- Variant · format · precision
- NVFP4 variant · safetensors · NVFP4
- Weights SHA-256
2228b9b22e93a88d84556419c879448ab6c490ae65c4c0b166f4962190ddbf26d3cfb92742e30c8b46564665791dbe0a86ed64cfc02b1275081530793c0c9581db6146a5464fb0a891181b93c81593f0ca65c602eb14120a1c2b1b09bca11f85fbcdb5ba1cdda462b5f38592d071e772c4d398afea61a0aa9188b32d1a239a79- Engine · version · feature
- SGLang ·
0.0.0.dev0+qwen38.27b.g561c8f3· 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
ae06db6e13f7e0df59a2831f51932b997a654b4e3402352238c514019f5351e2
Workload shape
- cache state
- cold
- People in this run
- 5
- Configured window (tokens)
- 262144
- Filled content target (tokens)
- 64000
- fixture variant
- distinct-v1
- KV cache precision
- fp8_e4m3
- Output token ceiling
- 512
- scenario
- distinct
- 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 | 248.80 s | machine total | single | not recorded | wall_seconds | |
ttft median s | 140.65 s | per person | median | 5 | ttft_median_s | |
hidden facts total | 15.00 count | machine total | single | not recorded | needle_total | |
ttft max s | 239.41 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 | 4.00 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 | 2.62 tok/s | machine total | single | not recorded | aggregate_tps | |
aggregate e2e tps | 2.13 tok/s | machine total | single | not recorded | aggregate_end_to_end_tps | |
peak host unavailable gib | 114.20 GiB | machine total | max | not recorded | peak_host_unavailable_gib |
Data and citation
Skilak Keep evidence, run 1281f46a33da47d2b824c004, measured 2026-10-06T11:32:14.407008+00:00. https://skilak.ai/keep/resources/evidence/r/1281f46a33da47d2b824c004
