HomeTechnology
May 26, 2026 · Evening edition
Huawei Technologies has announced a significant breakthrough with its new 'Tau Scaling Law,' a principle enabling the company to develop more advanced chips without requiring specialized U.S. equipment, which has been subject to blockades. This development, focusing on reducing data movement time within chips rather than merely shrinking transistor size, led to a notable increase in semiconductor company shares globally. Huawei anticipates designing high-end chips comparable to those manufactured with a 1.4-nanometer process by 2031.
Huawei Technologies’ claim of a new route to more advanced chipmaking sent semiconductor shares higher globally, adding momentum to a technology market already being reshaped by heavy artificial intelligence spending, falling model prices and rising security concerns.
The company said its new “Tau Scaling Law” could allow it to develop more advanced chips without relying on specialized U.S. equipment that has been subject to blockades. Rather than focusing only on shrinking transistor size, the approach emphasizes reducing the time required to move data within chips. Huawei said it expects the method to support designs for high-end chips comparable to those manufactured with a 1.4-nanometer process by 2031.
The announcement underscored the central pressure point in global technology: demand for AI computing is rising quickly, while the hardware supply chain remains constrained by manufacturing complexity and geopolitical limits on access to advanced equipment. It also highlighted why investors remain highly responsive to any sign that chipmakers or major technology companies may find new ways to improve performance outside the conventional path of process-node scaling.
That optimism is being tempered by financial-stability concerns. The European Central Bank warned that the rapidly expanding AI investment boom, increasingly funded by private credit, has become a new risk for the euro area’s financial system. The ECB said insurers, pension funds and other investors could face significant losses if AI technology does not live up to its high expectations.
Competitive pressure is also intensifying in the AI software market. DeepSeek made a 75% price cut on its V4-Pro generative AI model permanent, sharply lowering inference costs and escalating price competition among frontier-class models. The move is expected to put pressure on rivals including Anthropic and OpenAI to revisit their own pricing strategies.
Browse a few recent editions, or open the full archive.
At the same time, enterprises are facing a broader security challenge from more capable AI systems. BNP Paribas has partnered with Mistral to strengthen defenses against potential threats from models such as Anthropic’s Mythos, which has demonstrated an ability to identify numerous high-severity cybersecurity vulnerabilities. Separately, TELUS Digital released a GenAI Safety Model Benchmark based on more than 620,000 adversarial tests, warning that AI models can be coaxed into unsafe behavior and calling attention to the need for continuous, automated security testing backed by human oversight.
The hardware ecosystem is also adapting to AI’s demands beyond chip design itself. Advanced Semiconductor Engineering unveiled an automated 310mm by 310mm panel-level packaging production line, which it described as an industry first. The company said the larger panel format is designed to improve throughput, reduce cycle time and increase material efficiency for complex multi-die architectures used in AI accelerators and high-performance computing devices. The line is scheduled to enter production in the first half of 2027.
Together, the developments point to a technology sector moving on several fronts at once: alternative paths to advanced chips, larger bets on AI infrastructure, cheaper model inference and more urgent efforts to contain the risks created by increasingly powerful systems.
Huawei Technologies has announced what it describes as a significant advance in chip development, saying its new “Tau Scaling Law” could allow the company to build more advanced processors without relying on specialized…
Read full articleThe European Central Bank has warned that the fast-growing artificial intelligence investment boom is creating a new risk for the euro area’s financial system, particularly as more funding comes from private credit.
Read full articleDeepSeek has made a 75% price reduction on its V4-Pro generative AI model permanent, sharply lowering the cost of inference and escalating price competition in the artificial intelligence market.
Read full articleConcern is growing over the ability of advanced artificial intelligence models to uncover cybersecurity vulnerabilities, prompting new defensive efforts and calls for stronger testing of enterprise AI systems.
Read full articleAdvanced Semiconductor Engineering has unveiled what it calls an industry-first automated 310mm × 310mm panel-level packaging production line, aimed at accelerating development of AI and high-performance computing hardw…
Read full article