Semiconductor News & Analysis Feed

NVIDIA has released two open-source AI technologies—Nemotron 3.5 Lightning and NeMo Switchyard—designed for efficient execution in agentic AI workflows. Nemotron 3.5 Lightning is a 30B parameter Mixture-of-Experts model with only 3B active parameters, built on a hybrid Mamba-2 + MoE + Attention arch

NVIDIA's technical blog explores how advanced fusion kernels significantly boost the training throughput of Mixture-of-Experts (MoE) models. As MoE becomes a foundational component of large-scale AI systems, optimizing these blocks is critical for maximizing training efficiency. NVIDIA introduces cu