Chinese scientists have announced the development of a revolutionary analog computing chip that they say can outperform today’s most advanced digital graphics processing units (GPUs) by up to 1,000 times while consuming only a fraction of the energy.

The breakthrough, led by researchers at Peking University and published on October 13 in the journal Nature Electronics, represents what experts describe as a major leap in semiconductor technology. The chip, based on analog computing principles, performs calculations using physical electrical currents instead of the binary 0s and 1s used by traditional digital processors.

According to the research team, the new chip is built from arrays of resistive random-access memory (RRAM) cells. These cells can both store and process data by controlling how easily electricity flows through them. This allows the chip to compute directly within its hardware—eliminating the need to transfer data between processor and memory, one of the main energy bottlenecks in digital systems.

In tests, the analog chip successfully solved complex mathematical problems, including matrix inversion tasks used in large multiple-input multiple-output (MIMO) communication systems. The results matched the accuracy of conventional 32-bit digital processors while using about 100 times less energy.

When scaled to more demanding workloads, the chip’s performance exceeded that of top-end GPUs such as Nvidia’s H100 and AMD’s Vega 20, achieving throughput up to 1,000 times higher, according to the study’s findings.

Analog computing, though conceptually older than digital computing, has historically faced limitations in precision, scalability, and practicality. The Peking University team claims to have solved many of these issues by designing circuits capable of maintaining digital-level accuracy while operating at analog speeds.

“By integrating storage and computation in the same physical hardware, we break through both the data and energy bottlenecks that have limited digital systems for decades,” the researchers said in their paper.

The new chip could have major implications for artificial intelligence (AI), 6G communications, and edge computing, fields that demand ever-faster processing and lower power consumption. Conventional digital chips, including high-performance GPUs, are increasingly constrained by energy costs and data transfer inefficiencies.

Analysts note that this development aligns with China’s broader strategy to achieve semiconductor self-reliance and reduce dependence on Western chipmakers. Earlier this year, Beijing introduced new subsidies to boost domestic AI chip production and reduce reliance on foreign GPUs by 2027, according to a Reuters report.

Despite the promising results, experts caution that moving from laboratory prototypes to commercial production will require further validation. Large-scale manufacturing consistency, software integration, and real-world reliability remain open challenges.

If successful, however, the technology could reshape the global computing landscape—potentially offering a more energy-efficient and faster alternative to the current digital architectures that power AI systems worldwide. (ILKHA)