According to the latest data received from a well-known insider under the nickname kopite7kimi, the future flagship video card Nvidia GeForce RTX 5090 will be almost twice as powerful as its junior model RTX 5080. VideoCardz reports this.
Key specs for the RTX 5090 include a GB202-300-A1 GPU, 21,760 CUDA cores, 32GB of GDDR7 VRAM, a 512-bit bus, and 600W power consumption. In turn, the RTX 5080 can be equipped with a GB203-400-A1 GPU, 10,752 CUDA cores, 16 GB of GDDR7 video memory, a 256-bit bus and a power consumption of 400 W.
A comparison table published in VideoCardz shows a significant improvement in the performance of video cards from the RTX 40 and RTX 50 lines. If the information is confirmed, the flagship RTX 5090 will receive almost 30% more CUDA cores compared to the previous model, increased memory by 8 GB and expanded bus width from 384 to 512 bits. However, the power consumption of the video card will also increase significantly from 450 to 600 W.
The RTX 5080's improvements over the RTX 4080 will be less significant. The number of CUDA cores will increase from 9728 to 10,752, and the amount of memory will remain the same – 16 GB (256 bits), but the memory type will change from GDDR6X to GDDR7.
Experts from VideoCardz emphasize that the RTX 5090 will be almost twice as powerful as the RTX 5080, which is significantly larger than the performance gap between top-end video cards in the previous generation.
Sales of video cards from the RTX 50 line are expected in 2025. NVIDIA is expected to introduce graphics accelerators in the coming months, although the company itself has not yet made any official announcements related to next-generation graphics cards.
