Semiconductor News & Analysis Feed

Multiverse Computing and Qualcomm Technologies have partnered to deliver highly efficient AI models to data centers, leveraging Qualcomm's Dragonfly AI200 and AI250 accelerators and Multiverse's model optimization technology. This collaboration aims to reduce compute and memory footprints, enabling

Multiverse Computing launched its Pulsar 16B reasoning model on June 22, 2026, built on NVIDIA's Nemotron architecture. The model achieves performance comparable to 30B-class systems using only 16 billion parameters, leveraging Multiverse's CompactifAI technology to eliminate mathematical redundancy