The C++/SYCL-based Intel Arc inference engine machx-inference-engine released v0.2.35 in a GitHub commit on November 2, 2026, adding an N-stage chain that runs Flash-Next split across multiple cards, and fixing a community GGUF reading error.
The C++/SYCL-based Intel Arc inference engine machx-inference-engine released v0.2.35 in a GitHub commit on November 2, 2026, adding an N-stage chain that runs Flash-Next split across multiple cards, and fixing a community GGUF reading error.