Recently, a research team headed by Researcher Sun Zhong from the Institute for Artificial Intelligence at Peking University, in conjunction with a research team from the School of Integrated Circuits, published a paper in the internationally acclaimed academic journal Nature Electronics. The paper, titled "Precise and Scalable Analog Matrix Equation Solving Using Resistive Random-Access Memory Chips," represents a substantial leap forward in the realm of novel computing architectures.

The research team has successfully crafted a high-precision, scalable analog matrix computation chip that leverages resistive memory technology. This innovative chip marks the first instance where analog computing achieves a level of precision comparable to that of digital computing. By elevating the precision of traditional analog computing by an impressive five orders of magnitude, the chip attains a remarkable 24-bit fixed-point precision.

Performance assessments reveal that when tackling crucial scientific challenges, such as large-scale MIMO signal detection, the chip's computational throughput and energy efficiency far surpass those of the current top-tier digital processors (GPUs). Specifically, when computing the inverse of a 32×32 matrix, the chip's computational prowess outstrips the single-core performance of high-end GPUs. As the problem size escalates to a 128×128 matrix, the chip's computational throughput soars past that of top-tier digital processors by over 1,000 times.

Moreover, at an equivalent level of precision, the chip's energy efficiency is over 100 times greater than that of traditional digital processors. This groundbreaking achievement not only shatters the precision and scalability limitations that have long hindered analog computing but also paves entirely new avenues for tackling computational hurdles in domains such as artificial intelligence and 6G communications.