AI Briefing#metal
The latest AI and developer news about #metal, with the original source and a short summary.
The latest AI and developer news about #metal, with the original source and a short summary.
The release removes deprecated APIs and adds FP8 support, YaRN truncation controls, and performance improvements across CUDA, Metal, Vulkan, and other backends.
A new llama.cpp branch featuring Metal and n-gram SSD offload optimizations has been released, improving Qwen-3.8-Flash-Next inference performance on Apple Silicon.
Redis developer antirez released h3.c, a C/Metal-based inference engine that runs the MiniMax H3 model on Apple Silicon.
llama.cpp inference speed on Apple Silicon macOS VMs was increased up to 16x using a Metal compatibility layer.
The SDK 'Cider,' which supports W8A8 activation quantization on Apple Silicon by addressing MLX's limitations, has been released.
DwarfStar 4, a Metal-based local inference engine, has been released supporting DeepSeek V4 Flash.
ds4.c introduces a Metal-only inference engine and server built specifically for DeepSeek V4 Flash.
A hackathon was held where participants use OpenAI Codex to optimize Mac Metal kernels.
A Metal kernel implementation combining INT3 model compression with INT2 KV cache has been released.
By sharing Wasm memory with the GPU, inference runs zero-copy on Apple Silicon.