Niko Veit (@coldniko)
Feed do autor Strata.
Sync 2026-10-08 17:27 UTC
Posts
- @rskjr Hey Robert! I have seen your live stream and concerns in Strata and totally agree. Things are progressing a bit fast, just wanted to …2026-10-06 · update No X
- @yue_neg_631 I have a surprise for that :))2026-10-06 · update No X
- Strata wins by a tight margin against other Strix Halo inference engines on first try release
Community benchmark attention on Ryzen AI Max / Strix Halo builds with Strata v0.1.40.
- @servasyy_ai Amazing work from you! I love the community and all the help I get with great PR's and benchmarks you are doing. I'll be lookin…2026-10-06 · benchmark No X
- @toolowgear @__kutluk Soon 👀👀👀2026-10-06 · update No X
- Only loses to @ciruai engine on Hermes benchmarks. Kudos 🫡2026-10-06 · benchmark No X
- Strata wins by a tight margin against all other Strix Halo inference engines in it’s first try release. Did not expect that! 🥳 https://t.co…
- @joseed801 Yes2026-10-06 · update No X
- @badguyty @LotusDecoder I have Aura, but it’s catching up on me 😭2026-10-06 · update No X
- @Biggest I’m waiting @Alibaba_Qwen @QwenDevs 👀2026-10-06 · models No X
- @shishimaimidori You are correct and it packs better power.2026-10-06 · update No X
- @gabu3d_pl @pardoxAI I'm just doing tests on 4x cards. Hold up ;)2026-10-06 · update No X
- @D3PR3C4T0R @AMD @ColbyCakesss @AzorFrank I'm dealing with Strata, I have finger crossed for you to handle it!2026-10-06 · gpu No X
- @ibrahim_sanata5 Works steadily on 8GB2026-10-06 · update No X
- Running Strata on Mini PC with iGPU 780M — about 13 tokens/s on first try
APU / mini PC path for localized inference; pairs with install docs and model sizing.
- @RRunner144 Get me SSH!!!2026-10-06 · update No X
- @arnewc @forgalaforus There is a lot more to go.2026-10-06 · update No X
- @sami3dat Need to get W7800 to optimize -.-2026-10-06 · update No X
- @sami3dat Yes, but very soon optimized ;)2026-10-06 · update No X
- @zharoboy Check my latest post :))2026-10-06 · update No X
- @cryptodreki @LotusDecoder I will in 2-3 days tops2026-10-06 · update No X
- Running Strata on Mini PC with iGPU 780M2026-10-06 · hardware No X
- For those who slept on Qwen3.8-Flash-Next — community ts-bench comparisons vs 27B stacks
Highlights MoE local speed vs dense llama.cpp routes; links to models and install pages.
- Strix Halo officially supported in Strata v0.1.40
AMD GPU improvements, multi-GPU batching, decode and Turing/Arc fixes — see release notes on site.
Interações
- Community call to support Strix Halo / Strata hardware outreach — Ecosystem voices (e.g. Zed) amplifying AMD community hardware for Strata development. source