First reported by Tom's Hardware, the patch notes for the latest update to Nvidia's CUDA Toolkit state that support for the Maxwell and Pascall architectures⁠—GTX 9 and 10-series cards⁠—will be deprecated in an upcoming update.

Nvidia is launching the first volley of RTX 50-series GPUs based on its new Blackwell architecture, starting with the RTX 5090 and working downward from there. The company also appears to be winding down support for a few of its older GPU architectures, according to these CUDA release notes spotted by Tom's Hardware.

The RTX 5090 from Nvidia is not yet on sale, but the full version of the graphics processor is already appearing on the internet. With 800 watts of power dissipation and 24,576 shader units, the GPU is beyond good and evil.

NVIDIA's new GeForce RTX 5080 shows early benchmark results: 15% faster in synthetic 3DMark testing than the GeForce RTX 4080 SUPER.

We know that NVIDIA's GeForce RTX 5070 Ti and 5070 are expected to launch in February, we just don't know when exactly. Thanks to one major retailer in

Nvidia's GeForce RTX 5090 starts at $1,999 before you factor in upsells from the company's partners or price increases driven by scalpers and/or genuine demand. It costs more than my entire gaming PC.

NVIDIA's purported GeForce RTX 5090 Ti teased: huge 24576 CUDA cores, monster 800W power limit, 32GB of faster 32Gbps GDDR7 memory.

Nvidia hasn't announced the RTX 5060 yet, but I'm already worried about how the card will perform when it shows up.

With its AI capabilities enabled, the RTX 5090 is the fastest and best-performing graphics card in the world. Can Nvidia's competitors catch up?

Nvidia has revealed Maxwell, Pascal, and Volta architectures are considered feature-complete and will be frozen in an upcoming CUDA release.

That's a fairly hefty 13% bump on the 21,760 cores of the official RTX 5090. Prior to this alleged leak, the GB202 chip at the heart of the RTX 5090 was indeed thought to rock 24,576 cores. So, it all scans.