Whalesbook
Beyond Moore’s Law
For decades, the global semiconductor industry has followed Moore's Law, focusing on shrinking physical transistor sizes. Huawei Technologies, facing significant challenges due to its inclusion on the U.S. Entity List, is now trying a different path. At the 2026 IEEE International Symposium on Circuits and Systems, Huawei introduced the Tau Scaling Law. This new approach prioritizes optimizing time constants and reducing signal delays within electronic systems, rather than solely relying on physical transistor size.
Architectural Pivot: LogicFolding
This change is already being implemented. Huawei's upcoming Kirin chipsets, expected in late 2026, will feature an architecture called LogicFolding. This design aims to reduce internal wiring, thereby shortening signal paths and increasing system density without needing the advanced EUV lithography tools that Chinese manufacturers cannot access. Huawei states this method has been used in mass production for 381 chip designs over the last six years, covering mobile processors to AI accelerators.
Technical Risks Ahead
While Huawei's goal of reaching 1.4-nanometer equivalent density by 2031 is a significant strategic move, it faces considerable technical hurdles. Unlike industry leaders like TSMC and NVIDIA, which have access to global research, development, and materials, Huawei operates in a more confined environment. Limitations in accessing top-tier 3D NAND and HBM technologies remain a major challenge for its AI hardware. Additionally, although Huawei has improved AI storage with proprietary packaging, its memory technology could be vulnerable to future U.S. regulations targeting downstream memory components.
Market Context and Future
Industry experts are uncertain if internal design innovations can fully overcome the physical limitations of semiconductor fabrication. Chinese companies like Alibaba and ByteDance are adapting their systems to use Huawei's Ascend architecture to reduce reliance on U.S. chips. However, a performance gap persists between Huawei's offerings and the leading edge of global AI hardware. As of mid-2026, Huawei's strategy centers on achieving technical self-sufficiency. Its success will depend on whether these architectural advancements can keep pace with the rapid evolution of global AI inference and training demands.