[News] AI-Designed CPUs Match Decade-Old Smartphone Chips, Cambricon Says; AI Also Reshapes Materials, Equipment

AI accelerators are powering the AI boom, but the relationship is becoming a two-way street, as AI is increasingly reshaping the semiconductor industry. According to Shanghai Observer, citing Cambricon Technologies co-founder Chen Yunji, AI-designed chips have already reached performance levels comparable to human-designed processors from more than a decade ago.

AI-Designed Chips Make Rapid Strides

Chen made the remarks at Saturday’s 2026 World Laureates Forum in Shanghai. According to Shanghai Observer, he highlighted two major hurdles in AI-driven chip design: the vast number of possible CPU architectures and the near-perfect accuracy required for processors executing billions of operations per second.

To overcome these challenges, his team developed an iterative system that automatically verifies and corrects AI-generated designs hundreds of times per second, rather than relying on a flawless first attempt, the report suggests.

The approach produced Qimeng-1 in 2021, with performance comparable to Intel’s 486 processor from the 1990s, Chen said. By 2025, its successor, Qimeng-2, had reportedly reached performance levels similar to mainstream smartphone CPUs from more than a decade earlier.

Beyond AI-driven chip design, Chen’s team also analyzed more than 100 AI algorithms and identified a common set of underlying computational operations, paving the way for a dedicated AI instruction set architecture (ISA), according to Shanghai Observer. Huawei subsequently adopted the team’s processor architecture for its first-generation AI chips, paying approximately CNY 200 million in licensing fees, Chen noted.

AI-Driven Materials Design: The Apple Watch Case

Meanwhile, AI is extending its reach into materials science. According to China National Radio, Nobel laureate Konstantin Novoselov highlighted AI’s potential to modify the atomic structures of two-dimensional (2D) materials and predict the resulting properties. However, he also cautioned that real-world materials are far more complex than AI-generated models, which often rely on idealized substances such as perfectly pure iron or silicon.

Meanwhile, inverse design is gaining traction in materials development. Rather than starting with a material’s composition, researchers define target properties first, then work backward to identify suitable structures, compositions and manufacturing processes, the report notes.

Northwestern University professor Jeffrey Snyder cited the Apple Watch as an example. While its exterior may resemble conventional stainless steel or aluminum alloy, Snyder said it uses a proprietary alloy developed through inverse design, with properties such as strength, corrosion resistance, reliability, machinability and polishability tailored to specific requirements.

AI Reshapes Semiconductor Equipment

AI is also transforming semiconductor equipment, where complex processes make conventional physics-based tuning costly and time-consuming. According to YiCai Global, citing Piotech Chairman Lu Guangquan, equipment makers are increasingly turning to AI-driven virtual manufacturing to improve efficiency.

Lam Research, for instance, uses physics-based digital twins to accelerate certain etch-modeling steps by up to 30 times, while Applied Materials reportedly employs AI for real-time process analysis, handling more than one million 3D wafer measurements per hour.

China’s Piotech is pursuing similar advances, using bidirectional data mapping between physical and virtual process chambers for real-time process prediction, predictive maintenance and wafer transport optimization, the report notes, citing Lu. Lu also noted that AI agents can automate tasks such as failure analysis, cutting workloads that once required hundreds of labor-hours to under 5.5 hours.

Taken together, these developments underscore AI’s expanding role across the semiconductor value chain—from designing processors and engineering materials to optimizing manufacturing equipment. For China, advances in AI-driven design and manufacturing could offer additional avenues for innovation amid restrictions on access to advanced chipmaking technologies.

Read more

• [News] China AI Model Developers Ramp Up In-house Chips: DeepSeek, Zhipu Explore Custom Silicon
• [News] Huawei Says Ascend Market Share Tops NVIDIA in China, Backs NPO Over CPO for Optical Interconnects

(Photo credit: FREEPIK)