Following the US Supreme Court ruling that US President Donald Trump had acted unlawfully in invoking the International Emergency Economic Powers Act (IEEPA) to impose sweeping global tariffs, Trump swiftly announced a new across-the-board 15% tariff hike. Taiwanese steelmaker China Steel Corporation (CSC) has stated that the tariff dispute is far from over and may intensify further, requiring close monitoring of future developments.
Walsin Technology Corporation announced plans to invest JPY508 million (US$3.27 million) to increase its shareholding in Japan's Matsuo Electric, a move the passive components maker says will support product expansion and strengthen the two companies' partnership.
HP reported first-quarter results for fiscal year 2026 (ending January) that exceeded expectations, driven by the Windows 11 upgrade cycle and surging demand for AI-powered PCs. Yet the company cautioned that escalating memory prices and the US trade policy uncertainties could push full-year profits to the lower end of guidance, while overall PC shipments for FY26 are expected to decline in the double digits.
Chinese robotics firm Unitree Robotics unveiled its latest quadruped robot, the Unitree As2, on February 24, 2026, positioning the new model as a more capable platform for industrial and outdoor applications. The As2 builds on earlier consumer-oriented products and delivers a combination of performance upgrades and enhanced utility designed for real-world deployment.
A recent OpenAI disclosure has reignited debate about the company's AI infrastructure ambitions — but the alarm is largely misplaced. According to CNBC, OpenAI told investors its total compute spending target through 2030 would be approximately US$600 billion. That figure was quickly set against CEO Sam Altman's earlier pledge of US$1.4 trillion in infrastructure investment, leading some to conclude the company is pulling back sharply. It is not.
Advanced Micro Devices (AMD) and Meta have entered into a multi-year agreement to deploy up to 6 gigawatts of AMD Instinct GPUs to power Meta's next-generation AI data centers, marking one of the largest AI infrastructure partnerships to date and potentially exceeding US$100 billion in value.
Nvidia's high-performance GPUs sit at the heart of the US-China AI chip standoff — and Washington's export controls mean every policy shift lands directly on its order book. Two months after the Trump administration signaled it would allow limited H200 sales to China, not a single chip has been approved, a senior US export enforcement official has confirmed.
Texas Instruments (TI) has provided a revenue guidance range of US$4.32 billion to US$4.68 billion for the first quarter of 2026. This outlook reflects significantly stronger seasonality than usual, with revenue typically ranging from a low-single-digit decline to flat. Analysts noted that this sequential growth guidance marks what appears to be the first of its kind for the company in roughly 15 years. Earnings per share for the same period are expected to range from US$1.22 to US$1.48.
Widespread speculation surrounds Nvidia CEO Jensen Huang's previously teased chips set to be revealed at the March 2026 GTC conference. Among the most discussed possibilities are the long-rumored N1 and N1X PC chips, which have circulated in industry reports for the past two years but have yet to be officially announced. The N1 series chips appear ready and have reportedly completed product development with major PC brands, with Lenovo said to be one of the most critical partners in the collaboration.
Apple is expected to hold a product launch event in March 2026, with a budget MacBook anticipated to take center stage. Supply chain sources indicate volume shipments are projected to begin that same month, with some key components already delivered since late 2025. Rising prices for critical components — memory in particular — are making it harder to keep the model affordable, and could weigh on shipment volumes.
Toronto-based AI chip startup Taalas says it can hardwire a large language model directly into silicon to accelerate inference beyond what conventional GPUs can deliver. Founded in 2023, its first product — the HC1 inference chip — generates nearly 17,000 tokens per second for a single user running Meta's Llama 3.1 8B. In company benchmarks, performance is reported at 48 times that of Nvidia's B200 under the same configuration.
More coverage
