Lese-Ansicht

Chinese enthusiast overclocks AMD's Ryzen 7 9800X3D to 7.33GHz on Colorful X870E Vulcan motherboard — New world record achieved using liquid nitrogen cooling

A new record-high clock speed has been achieved on AMD's Ryzen 7 9800X3D. The best gaming CPU in the world doesn't need to be overclocked to achieve exceptional performance, but Chinese overclocker "Hero" has just pushed it to 7,335 MHz using a Colorful iGame X870E Vulcan motherboard, using liquid nitrogen.

  •  

U.S. allows TSMC to import chipmaking equipment to its China fabs — Samsung, SK hynix likewise receive go signal from Commerce Department

The U.S. Department of Commerce didn't renew the validated end-user status of these chipmakers, requiring them to acquire annual licenses to import chipmaking tools containing U.S. tech into their Chinese fabs.

  •  

ASRock to enter AIO liquid cooling market with an intriguing pumpless design at CES — full lineup to include more traditional options, some with LCD displays

ASRock, usually known for producing motherboards, is set to debut its own AIOs at CES 2026, including flagship models featuring LCD screens. It's unclear if ASRock is licensing Asetek's pump designs, and whether it will compete against affordable brands like Thermalright.

  •  

UK company shoots a 1,000-degree furnace into space to study off-world chip manufacturing — semiconductors made in space could be 'up to 4,000 times purer' than Earthly equivalents

British startup Space Forge has just successfully turned on its first in-orbit furnace, reaching 1,000 degrees Celsius in space. This is a crucial first step for the company in studying the feasibility of manufacturing chips in space.

  •  

Musk to expand xAI's training capacity to a monstrous 2 gigawatts with third building at Memphis site — announcement comes days after Musk vows to have 'more AI compute than everyone else'

Elon Musk has announced that xAI has purchased a third building at its Memphis, Tennessee site to bolster the company's overall compute power to a gargantuan two gigawatts.

  •  

Nvidia's CUDA Tile examined: AI giant releases programming style for Rubin, Feynman, and beyond — tensor-native execution model lays the foundation for Blackwell and beyond

Nvidia's CUDA 13.1 introduces CUDA Tile, a new tile-centric programming path that elevates GPU kernel development above SIMT. The innovation aligns CUDA with the tensor-native execution model of Blackwell-class GPUs, and lays the software foundation for future architectures built around increasingly specialized compute and data-movement engines rather than thread-level parallelism.

  •  

TikTok owner ByteDance to reportedly purchase $14 billion worth of Nvidia AI GPUs in 2026 — Company betting on Beijing's approval following Trump admin's ease on AI export controls

ByteDance is reportedly planning to spend 100 billion yuan, or $14 billion USD, on Nvidia's AI GPUs in 2026. Specifically, Bytedance is eyeing up Nvidia H200 chips, following Washington's announcement that it will allow them to be sold to approved parties in China. Beijing still has to approve these transactions, however.

  •