Anton Shilov

TSMC is prepping a 9.5-reticle, 7,885 mm² multi-chiplet packages on 120×150 mm substrates with integrated power management for future AI and HPC processors.

TSMC first GAA-based N2 process will enter HVM in the second half of 2025 with strong early adoption from both mobile and HPC/AI sectors.

TSMC has begun production of chips on its performance-enhanced N3P node, plans to follow with the high-frequency N3X variant.

TSMC unveils its A14 process technology featuring 2nd generation GAA nanosheet transistors and NanoFlex Pro DTCO with production slated for 2028.

Russia aims to begin local mass production of Elbrus-based CPUs using domestic 28nm-class process technology by 2030, despite sanctions, tool limitations, and talent shortages.

Nvidia Blackwell AI datacenters get cooling upgrade

Intel is reportedly planning to cut over 20,000 jobs as part of new CEO Lip-Bu Tan's push to streamline management, reduce costs, and refocus the company amid ongoing business struggles.

More cash for more GPUs.

In its annual report, TSMC acknowledged it cannot fully track how its chips are ultimately used, highlighting ongoing risks of unintentional export control violations.

GAA transistors and backside power delivery seemingly do wonders.

Intel's Lip-Bu Tan has overhauled leadership by naming Sachin Katti as CTO and AI chief, flattening the structure, and taking direct control of key technology developments.

Huawei unveils AI CloudMatrix 384 system based on 384 Ascend 910C processors that can beat Nvidia's GB200 NVL72 in performance, but at the cost of efficiency.

Nvidia CEO reaffirmed the company's commitment to serving China despite escalating U.S. export restrictions.

The House Select Committee on the CCP seeks to know all of Nvidia's significant clients in Asia.

Data center hub or export control avoidance?

TSMC plans to produce 30% of its 2nm and 1.6nm chips in Arizona as part of a its effort to transform its Fab 21 site into a major U.S. semiconductor hub.

Delay, but not stop.

Apparently, Jensen Huang's $1 million dinner with Donald Trump did not help to ease U.S. government's concerns over H20 GPUs exports to China.

The scale and timeline of Nvidia's plans to build AI server supply chain in the U.S. look ambitious, but can they actually be fulfilled?

A Chinese blogger claims that Qualcomm's second-gen Snapdragon X processors for PCs featuring Nuvia-designed Oryon cores will start at 4.4 GHz and deliver 18–22% higher performance.

Tech could lower manufacturing costs by simplifying one of the most complex steps in chip production and 3D chip integration.

AMD tapes out and powers on a 2nm CCD for its 6th Generation EPYC 'Venice' processor.

Intel is selling a 51% stake in Altera to Silver Lake for $4.46 billion as part of a strategy to streamline operations and focus on core products.

To avoid looming U.S. tariffs and strengthen supply chain resilience, Nvidia is partnering with TSMC, Foxconn, Wistron, Amkor, and SPIL to localize production of its AI servers and chips in the U.S.

President Trump announced new semiconductor tariffs to be introduced next week as part of an effort to shift tech manufacturing to the U.S.