Issues

• Community benchmark: RTX 6000D 84G, unsloth-ud-q4_k_xl, Strata 0.1.43

  Niko1221/Strata#2108 · Filter by author ditupao opened on Oct 12, 2026

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

  Niko1221/Strata#2107 · Filter by author ditupao opened on Oct 12, 2026

• Community benchmark : 7900 XTX, D4-2666, swift-1.5-iq2_xs, Strata 0.1.43

  Niko1221/Strata#2104 · Filter by author ditupao opened on Oct 12, 2026

• Community benchmark on NVIDIA A10

  Niko1221/Strata#2102 · Filter by author chenwei0930 opened on Oct 12, 2026

• [Linux, HIP gfx1100] 0.1.42+ PLE probe auto-switches to --ple-io mmap on a loop-device disk: decode -20-25%, prefill unaffected; explicit --ple-io direct restores 0.1.41 speed

  Niko1221/Strata#2101 · Filter by author JC-S opened on Oct 12, 2026

• Sub-agent fan-out: 0% sibling prefix reuse without --prompt-cache-tail; 93% and 11.7x prefill with it (GTX 1080 Ti, IQ3_S, 0.1.43)

  Niko1221/Strata#2099 · Filter by author MrKop-git opened on Oct 12, 2026
  3

• Community benchmark: RTX 3090, Qwen3.8 Flash-Next UD-IQ4_XS, Strata 0.1.43

  Niko1221/Strata#2097 · Filter by author QuantBox723 opened on Oct 12, 2026

• "--prefill", "auto" with value 1,024 resulted in significant t/sec drop

  Niko1221/Strata#2096 · Filter by author Tachbek opened on Oct 12, 2026

• Community benchmark: RTX 3090, Qwen3.8 Flash-Next IQ3_S, Strata 0.1.43

  Niko1221/Strata#2095 · Filter by author QuantBox723 opened on Oct 12, 2026

• [0.1.43] One GPU, default flags: "lending the prompt path its slots failed: ... do not fit" when a request's chunk is smaller than the auto chunk, engine exits (code 1)

  Niko1221/Strata#2090 · Filter by author K8nsch opened on Oct 12, 2026

• strata_bench.py + --expert-profile-save: the bench run overwrites your calibrated live profile with what it learned

  Niko1221/Strata#2088 · Filter by author mmbeanz opened on Oct 12, 2026
  1

• Community Bench: 2026-10-11-community-rtx-5060ti-3060-split — RTX 5060 Ti 16 GB + RTX 3060 12 GB, unequal links, --layer-split planner + the two new 0.1.43 defaults

  Niko1221/Strata#2085 · Filter by author mmbeanz opened on Oct 12, 2026

• System hard-freezes during long-prompt prefill on NVIDIA Pascal = driver bug 6869496 (PMU halt from PCIe link retraining) — verified workaround: NVreg RmDisableGen2LinkRetraining=1

  Niko1221/Strata#2084 · Filter by author MrKop-git opened on Oct 12, 2026
  1

• 0.1.42 -> 0.1.43: decode -3.4% on RTX 4090 48 GB, IQ3_S, --resident-experts (MTP accept unchanged, prefill flat)

  Niko1221/Strata#2074 · Filter by author Tlyer233 opened on Oct 12, 2026

• [Feature]: GLM-5.3-Flash (glm5next) support — the glm53f fork proves it works in Strata; AMD/HIP would make it unique

  Niko1221/Strata#2067 · Filter by author artfix opened on Oct 12, 2026

• Engine hangs at layer 48 during decode when VRAM headroom drops below ~500 MiB

  Niko1221/Strata#2065 · Filter by author bang-imvt opened on Oct 12, 2026

• strata_bench.py mislabels mixed-GPU hosts and single-GPU runs on multi-GPU hosts

  Niko1221/Strata#2062 · Filter by author MinhHaDuong opened on Oct 12, 2026
  1

• # Community benchmark on 890M - Minisforum X1 Pro Ryzen AI 9 HX 470/96GB/Linux

  Niko1221/Strata#2061 · Filter by author htbmixbox opened on Oct 12, 2026

• gfx1030 / Windows: prompts read 16–20% slower since 0.1.40.2 — bisected to ac52c2c7 (#1240)

  Niko1221/Strata#2055 · Filter by author did-technomancer opened on Oct 12, 2026

• update.sh: failed engine build leaves BUILD.json stamped with new source hash — engine never rebuilt (stale binary reports old version)

  Niko1221/Strata#2054 · Filter by author MrKop-git opened on Oct 12, 2026
  1

• Default multi-GPU recommendation not ideal for asymmetrical GPUs

  Niko1221/Strata#2052 · Filter by author MaKla89 opened on Oct 12, 2026

• t

  Niko1221/Strata#2051 · Filter by author Sahil0015-btech opened on Oct 12, 2026

• Deterministic deadlock at the completion handshake after the last block: host at layer step 48 with the GPU already rang - 5/5 stalls identical, on 0.1.41 and 0.1.43 (RTX 4080 / Windows / IQ3_XXS native pack / 256K)

  Niko1221/Strata#2049 · Filter by author gaoyonghui123 opened on Oct 12, 2026

• 2026-10-12-community-rtx-3060-qwen3-8-flash-next-iq3-s

  Niko1221/Strata#2046 · Filter by author Tee8088 opened on Oct 12, 2026

• Community Bench: 2026-10-12-community-2x-2080ti-22gb-qwen3-8-flash-next-gsq-rco-iq3 (Proxmox 9 VM)

  Niko1221/Strata#2044 · Filter by author sense1024 opened on Oct 12, 2026