Ashley-PC hardware/software capture, 3 Oct 26 (via SSH inventory)

GPU:      NVIDIA GeForce RTX 2080 Ti, 11 GB (Turing, sm_75)
Driver:   32.0.16.1047 (591.47)
CPU:      Intel Core i9-9900KF @ 3.60 GHz (8C/16T, 2019)
RAM:      64 GB DDR4 (68,719,476,736 bytes reported)
OS:       Windows 11 (10.0.26200.0)
Python:   3.12 (CPython, user-site install)
Engine:   Strata 0.1.38 release build, engine/BUILD.json:
          archs [75, 86, 89, 120], ptx true, CUDA 13.0, vision gpu, portable
Model:    Qwen3.8-Flash-Next-GSQ-RCO-IQ3_XXS (47.0 GB pack + PLE split)
Serve:    strata-iq3_xxs.json — expert-cache auto, prefill auto, spec 4,
          MTP rt, max-context 131072, kv int8, kv-resident 32768, vision,
          vram-reserve 700, spec-min-p 0.5
Note:     engine bound 127.0.0.1:8080; measured from a second machine
          (big-office, Linux) through an SSH tunnel, ~2 ms LAN RTT.
          API-side and engine-side numbers agree within ~1%.
