A web developer going by "@eisneim" on X (formerly Twitter) claims to have seen an RTX 4090 with 96GB of VRAM, quadruple the 24GB normally available

Infrastructure providers have built robust foundations offering GPU computing resources. However, the cyclical nature of GPU demand combined with substantial hardware investments creates profitability challenges. By expanding vertically into full-stack services, providers can boost GPU usage while accessing higher-margin revenue streams.

GPU shipments grew by 6.2% late last year despite shortages, but researchers predict that the graphics card market will decrease by 1% annually from 2024 to 2028.

JPR notes worldwide GPU shipments increased 4.4% quarter-over-quarter in Q4 2024 to 76.9 million shipments, data center GPUs saw massive growth.

As Nvidia seems to fumble the launch of the RTX 50-series, we wonder whether this is just a slow start or if the entire generation has gone to waste.

CoreWeave, the upstart GPU cluster datacenter operator that was formerly a relatively small cryptocurrency miner based in Roseland, New Jersey has filed

Asus reveals Nvidia RTX 5070 launch pricing, and it looks steep in places – complete with hints that GPU stock could be shaky, too.

NVIDIA CEO Jensen Huang and industry veterans will present 'what's next in AI' at GTC 2025, expect Blackwell Ultra and Rubin AI GPU tease (hopefully).

Nvidia is well-positioned to dominate the AI and data center GPU market, driven by strong capex investments from key cloud customers like Microsoft,

With all eyes on Nvidia, here's a look at the company's recent earnings trends. Read about the chipmaker's revenues, profits, and stock price moves.