by 
Nvidia's flagship RTX 5090 graphics card is impossible to buy anywhere near its original price, as its excellent compute performance and VRAM allocation has seen it become a tool for AI workloads instead. Now, Nvidia is reportedly formalising that shift in roles, with the GB202 GPU dies being allocated to RTX Pro workstation cards and the RTX 5090 being effectively discontinued.
The news comes from prolific leaker MEGAsizeGPU, who we last mentioned when rounding up next-gen GPU rumours, but has not been confirmed by Nvidia.
If RTX 5090 and related chips like the China-specific RTX 5090 D become actually unobtainable, rather than merely eye-wateringly expensive, then Nvidia's top-end GeForce card would become the RTX 5080 16GB.
Interestingly, there is a separate rumour from another leaker, hongxing2020, that states that the RTX 5080 will get a 24GB variant using 3GB GDDR7 chips - perhaps along similar lines to the rumoured RTX 50-Series Super lineup, which is looking less likely at this stage.
We don't expect to see proper next-generation GPUs from Nvidia (Rubin), Intel (Druid) or AMD (RDNA 5) until 2027 or even 2028, which is a similar window to next-gen consoles from Sony and Microsoft. However, the continuing AI-fuelled price surge in flash memory components has already caused significant disruption to everyone's plans, and looks set to continue for the foreseeable future.