Browsing: LLM inference

SiMa.ai has launched the Modalix PCIe HHHL Card, a compact and power-efficient AI accelerator built for Physical AI and LLM workloads at the industrial edge. Developed with Advantech, the card doubles performance over its predecessor and supports multimodal models, 16 video channels, GMSL camera connectivity, and full MLSoC processing—bringing real-time vision reasoning to legacy and modern industrial PCs without system redesign …

Read More