Issues

• [Engine] track explicit Vulkan product feature coverage

  Zutfen-LLC/Strata#70 · Filter by author ezutfen opened on Oct 28, 2026

• [Engine] add native-style portable prefill chunks and skip unused head work

  Zutfen-LLC/Strata#69 · Filter by author ezutfen opened on Oct 28, 2026

• [Engine] follow native shared and routed expert scheduling in Vulkan

  Zutfen-LLC/Strata#68 · Filter by author ezutfen opened on Oct 28, 2026

• [Engine] retain QSA attention execution and device KV state

  Zutfen-LLC/Strata#67 · Filter by author ezutfen opened on Oct 28, 2026

• [Engine] track native CUDA and portable Vulkan product parity

  Zutfen-LLC/Strata#65 · Filter by author ezutfen opened on Oct 28, 2026

• [Engine] add opt-in persistent mutual TLS for native GPU workers

  Zutfen-LLC/Strata#26 · Filter by author ezutfen opened on Oct 26, 2026
  6

• [Engine] prove an exact leading-prefix fold without regrouping

  Zutfen-LLC/Strata#21 · Filter by author ezutfen opened on Oct 25, 2026

• [Benchmark] assess optional gigabit transport tuning

  Zutfen-LLC/Strata#19 · Filter by author ezutfen opened on Oct 25, 2026

• [Benchmark] evaluate locality-aware placement and broader CUDA fanout on gigabit

  Zutfen-LLC/Strata#18 · Filter by author ezutfen opened on Oct 25, 2026
  6

• [Benchmark] bound persistent or chunked compression benefits

  Zutfen-LLC/Strata#17 · Filter by author ezutfen opened on Oct 25, 2026

• [Benchmark] track exact inference latency and wire volume on gigabit

  Zutfen-LLC/Strata#11 · Filter by author ezutfen opened on Oct 25, 2026

• [Engine] overlap CPU weight reads with bounded parallel verification

  Zutfen-LLC/Strata#4 · Filter by author ezutfen opened on Oct 23, 2026