Latest about GPUs
Nvidia says H20 export controls didn’t stop China’s AI progress
Nvidia shared an opinion piece by Aaron Ginn, co-founder of AI company Hydra Host, stating that despite the U.S.’s export controls on Nvidia’s H20 chips, China continued to achieve AI breakthroughs.
Is 8GB of VRAM enough for gaming? GeForce RTX 5060 Ti 8GB vs RTX 5060 Ti 16GB GPU face-off
Nvidia's GeForce RTX 5060 Ti comes paired with 8GB and 16GB of memory. We dive deep into our detailed benchmark results to see which one wins in a head-to-head battle.
Early RTX 5090D V2 benchmarks suggest that maybe 384-bit memory is just fine after all
There's barely any difference in performance between the RTX 5090D and its 512-bit interface compared to the 5090D V2 and its skinnier 384-bit interface in games.
AMD's elusive Radeon AI Pro R9700 makes its first retail appearance for the DIY market
It's time to ROCm.
Intel's new "Shared GPU Memory Override" feature allows you to manually allocate VRAM on some Core Ultra iGPUs
You control your destiny.
I paired a cheap $125 PSU with a $3,000 RTX 5090
The RTX 5090's 575W TGP is intimidating, but some cheap PSUs are still up to the task
The tale of Nvidia's HGX H20: How an AI GPU became a political lightning rod
Premium Nvidia’s HGX H20, a cut-down AI GPU originally allowed for sale to China, has become a geopolitical bargaining tool as the U.S. taxes its exports and China halts purchases over security concerns.
Intel releases LLM Scaler 1.0 for Project Battlematrix — Arc Pro B-series GPUs get enablement support and performance optimizations
Project Battlematrix is almost ready for deployment
