At the 2026 World Artificial Intelligence Conference (WAIC 2026), Qimor (奇异摩尔) unveiled its full-stack, domestically produced ultra-node interconnection solution. This solution aims to address the "single point strong, system weak" structural imbalance in China's AI computing power by enabling system-level synergy. A key demonstration was the IBGDA solution, developed with Biren Technology (壁仞科技), which allows domestic GPUs to directly communicate with domestic RDMA network cards, significantly boosting throughput and reducing latency for AI model training and inference by bypassing the CPU. AI
IMPACT This development could accelerate the adoption of domestically produced AI hardware by improving the performance and efficiency of AI computing clusters.
RANK_REASON This cluster details a significant advancement in domestic AI infrastructure, focusing on a new technological solution for GPU-RDMA communication that aims to reduce reliance on foreign components and improve performance. [lever_c_demoted from significant: ic=1 ai=0.7]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
