OpenAI on June 24 unveiled "Jalapeño," its first in-house AI accelerator, co-developed with Broadcom and manufactured by TSMC — the clearest sign yet that leading model developers want to design their own chips and broaden a supply chain long dominated by Nvidia.
Qualcomm used its June 24 investor day to unveil a data center chip portfolio and a multi-generation CPU agreement with Meta, a strategic bid to diversify beyond a shrinking smartphone business and challenge Nvidia's dominance — while courting Chinese demand and leaning on manufacturing partner TSMC.
On June 24, Micron Technology reported record fiscal third-quarter results and issued a fourth-quarter forecast that significantly exceeded market expectations, reinforcing investor confidence in the AI infrastructure boom. Analysts said the company's expanding portfolio of multi-year customer agreements could help reduce the memory industry's traditional volatility, while persistent supply constraints are expected to support pricing and profitability through at least 2027.
National Silicon Industry Group, China's leading 12-inch silicon wafer maker, has announced a CNY11.45 billion (US$1.6 billion) capital injection into its core subsidiary Shanghai Xinsheng Semiconductor Technology, in a restructuring aimed at easing losses in its wafer business.
Phison Electronics CEO Khein-seng Pua said the NAND flash market is facing a supply crunch with "no end in sight," as AI continues to crowd out memory capacity and push storage demand higher.
Nvidia and Amazon Web Services (AWS) are expanding tools that could make it easier for companies worldwide to build and run large-scale AI systems. The changes aim to improve speed, lower costs, and reduce operational complexity across inference, search, and training, which could influence how global enterprises deploy production AI.
SK Group chairman Chey Tae-won is planning to meet Tesla and SpaceX leadership in the US at the end of June 2026 to advance cooperation on next-generation AI infrastructure, memory supply, and data-center projects, South Korean outlet Ddaily reported, citing industry sources. The meeting is expected to cover specific business collaboration plans with Tesla, SpaceX, and xAI, though the exact date was kept confidential.
Techzone Technology Materials, a provider of one of Taiwan's few integrated waste treatment, renewable energy, and recycling solutions, is actively addressing the waste treatment demand driven by semiconductor capacity expansion. Together with its subsidiaries Han-Yang Technology Renewable Energy and Sunlight Environment, the company has a combined annual treatment capacity of 239,000 tons.
The market for fan-out panel-level packaging(FOPLP) and glass substrate packaging is set to grow sharply as chipmakers race to support AI and high-performance computing. For global readers, the shift could shape future device performance, supply chains, and where advanced semiconductors are manufactured and assembled.
China's LineShine supercomputer debuted at No. 1 on the June 2026 TOP500 list, announced at the ISC 2026 conference in Hamburg, becoming the first system to sustain more than two exaflops on the standard HPL benchmark using CPUs only. The result marks the first time since 2017 that a China-based system has led the TOP500 ranking, and reflects Beijing's effort to present a frontier computing system built around domestic processors, interconnects, and software.
IBM is building a foundry to produce silicon wafers used in quantum-computing processors, seeking to become an indispensable part of the quantum economy under development. Called Anderon, the independent subsidiary is set to begin production this year, a decade after IBM began exploring applications of quantum computing.
More coverage
