In 2026, driven by the explosive growth in demand for artificial intelligence and high-performance computing, the internal data transfer speed and efficiency of data centers have become a critical bottleneck limiting the full utilization of overall computing power