Daily AI news, sourcedUpdates / Chips & infra
Updates Chips & infra · Official · 44 items · newest issue first · page 1 of 3
AreaAllModelsAgents & codingApps & productsChips & infraResearchPricing & limitsBusiness & fundingPolicy & safetyOther
- AMD releases GAIA 0.25, its local AI agent app for Ryzen AI PCs, with a larger default model and a much smaller install
- Meta details Prometheus, the gigawatt-scale Ohio supercluster that trained Muse Wonder at more than 85% goodputOfficialEngineering at MetaOct 20Chips & infra
- doubleAI says its Maxwell code execution model made LTX-2.5 generate 4K video nearly 2x faster
- Hugging Face libraries now run natively on AWS Trainium through TorchNeuron, which enters public beta Oct. 20
- Kapitsa, an open-source inference engine for Apple Silicon Macs from Nace.ai developers, supports only Qwen and Nemotron modelsOfficialKapitsa (GitHub)Oct 19Chips & infra
- SGLang 0.5.22 speeds up DeepSeek-V4.1 serving by up to 2.3x on H200, adds a 10x faster model loader and runs on Apple Silicon
- vLLM 0.32.0 adds a structured-decision endpoint, NVFP4 KV cache on more GPUs and makes Model Runner V2 the ROCm default
- Ollama v0.40.7-rc0 adds PPLX Decider v1.2 support to its MLX backend and lets Responses requests set effort on non-thinking models
- llama.cpp merges faster CUDA matrix-vector kernels for low-bit quantizations, speeding up Q2_K and Q3_K generation on older GPUs
- Ollama v0.40.6 adds the q2_0 GGUF tensor type and fixes orphaned runner processes on Windows
- 4Paradigm open-sources ModelSphere, a production LLM deployment platform that schedules across Chinese AI chips
- PaddlePaddle releases FlashMask as a standalone attention library for long-context training, with PyTorch support
- Intel LLM-Scaler-vLLM 1.5 moves to vLLM 0.26 and adds Qwen3.8, Gemma 4, FP8 and MTP support on Intel GPUs
- Artificial Analysis posts first AA-AgentPerf results for NVIDIA Vera Rubin NVL72, up to ~5.7x GB300 throughput per MWOfficialArtificial AnalysisOct 16Chips & infra
- Codex Day 11 cuts 'Selected model is at capacity' interruptions and auto-continues when they still occurOfficialOpenAI (Codex)Oct 16Chips & infra
- Local Inference Lab publishes a ~98 GiB QAD NVFP4 build of Qwen3.8-Flash-Next that is smaller than NVIDIA's 133GB NVFP4 checkpointOfficialLocal Inference LabOct 16Chips & infra
- Microsoft adds experimental llama.cpp support to Windows ML so apps can run GGUF models locally through new task-specific APIsOfficialMicrosoft Foundry on Windows BlogOct 16Chips & infra
- Apple shows a new MLX release running Muse Spark Open at over 36 tokens per second across four M5 Ultra Mac Studios linked by RDMA over ThunderboltOfficialApple Machine Learning ResearchOct 16Chips & infra
- llama.cpp v0.7.0 moves the server's default port to 9931 and adds LiliCorr speculative decoding, K2 Horizon support and a web UI models manager
- Ollama v0.40.5-rc1 ports MLX Metal custom kernels to CUDA and accepts custom tool calls in Responses history