DeepSeek ports high-performance GEMM library to Huawei Ascend NPUs
DeepGEMM-Ascend brings API-compatible, near-peak matrix multiplication performance to Huawei Ascend 950 hardware, supporting FP4, FP8, and BF16 workloads.
Open-weight models, on-device AI and running models on hardware you control.
DeepGEMM-Ascend brings API-compatible, near-peak matrix multiplication performance to Huawei Ascend 950 hardware, supporting FP4, FP8, and BF16 workloads.
Hugging Face releases Olmo-core 3, an open infrastructure that boosts MoE training throughput and supports models with over one trillion total parameters.
Bespoke Labs has released Nimble, an open-weight 9B decision model fine-tuned from Qwen3.5-9B that provides probabilistic answers to structured questions in under 100ms on local hardware.
Magnitude compiles and tunes kernels on-device to run open models up to 2x faster than llama.cpp across Apple Silicon, NVIDIA, AMD, and CPU setups.
Google's 2026 guide details using LiteRT to run Gemma models on Raspberry Pi 5, achieving real-time inference for robotics and edge agents without cloud dependency.