Every Moment Deep Thinking Showcases Spatial Optical Computing at WAIC, Partners with Kstar Data
During the 2026 World Artificial Intelligence Conference (WAIC 2026), Every Moment Deep Thinking (每刻深思), led by its founder and chairman Zou Tianqi (邹天琦), showcased its spatial optical computing technology. The company exhibited its core technological achievements in this field alongside Jueyan (阶跃), presenting a full-stack integrated solution of spatial optical computing chips and large models. This solution demonstrated real-world reasoning acceleration using the Jueyan STEP 3.7 large model, aiming to reduce costs, improve efficiency, and enhance computing power performance. The exhibition also marked a strategic partnership between Every Moment Deep Thinking and Kstar Data (科华数据). This collaboration focuses on the synergistic integration of computing power and energy, specifically in areas such as Uninterruptible Power Supply (UPS) systems, liquid cooling, and heterogeneous computing power coordination. The partnership aims to accelerate the large-scale implementation of green computing solutions.
The integration of spatial optical computing with large language models, as demonstrated by Every Moment Deep Thinking at WAIC 2026, signifies a critical advancement in addressing the escalating demands for efficient AI computation. The partnership with Kstar Data to explore "computing-power-electricity-cooling synergy" highlights a systemic approach to optimizing data center infrastructure, crucial for sustainable AI development. As AI models grow in complexity and scale, innovations that reduce energy consumption and operational costs, such as optical computing and advanced cooling, will become increasingly vital. This trend points towards a future where specialized hardware and integrated infrastructure solutions are paramount for unlocking the full potential of AI while mitigating its environmental and economic footprint.
AI-generated to prompt reflection — not editorial opinion, not advice, not a statement of fact. How this works.