NVIDIA's GeForce RTX 5090 will have a giant graphics chip |
|
 |
|
|
On the Chinese forum Chiphell appeared a photo, presumably, of the PCB from the new flagship NVIDIA GeForce RTX 5090 graphics card. Apparently, it will get a giant GPU  GPU-accelerated computing is the use of a graphics processing unit (GPU) together with a CPU to accelerate deep learning, analytics, and engineering applications. and an impressive amount of memory.In the center of the board there is a place for the GB202 GPU GPU-accelerated computing is the use of a graphics processing unit (GPU) together with a CPU to accelerate deep learning, analytics, and engineering applications. with 21,760 CUDA cores and a frequency of 3 GHz (not including overclocking). It is claimed that the dimensions of the chip itself are 23x31 mm and 63x56 mm including the packaging. Around it there should be more than 40 capacitors and 16 GDDR7 memory modules of 2 GB each, which together gives 32 GB with 512-bit interface and bandwidth up to 1792 GB/s.Power is supplied to the board via a single 16-pin 12V-2x6 connector capable of handling up to 600W of power. Early tests have shown that it is not prone to overheating and melting like the 12VHPWR in the 40-series. The new NVIDIA GeForce RTX 50-series graphics cards are expected to be unveiled in two weeks, and will go on sale in January-March. The RTX 5060 models will be delayed and unveiled at a later date. |
