Nvidia’s Flagship RTX 5090 GPU May Ship As much as 46 % Enchancment

Nvidia GeForce RTX 50 Collection Blackwell GPUs had been introduced on the Client Electronics Present (CES) 2025 earlier this month however are but to hit the market. The RTX 5090 and RTX 5080 fashions might be out there for buy on January 30. Forward of their arrival, reviewers have reportedly revealed the artificial benchmarks for the top-of-the-line GeForce RTX 5090 GPU which trace in direction of a efficiency enchancment of as much as 46 %, relying on the benchmarking platform used to run the assessments.

Nvidia GeForce RTX 5090 GPU Benchmarks

In a report, VideoCardz compiled the benchmark take a look at outcomes. Citing conversations with a number of reviewers, the publication stories that players can count on a efficiency bounce of roughly 20 % with the brand new Nvidia GeForce RTX 5090 GPU in comparison with its predecessor, the RTX 4090. Whereas these numbers are pretty correct, the report emphasises that efficiency could range considerably relying on the sport, decision, and settings.

FireStrike, which is the DirectX 11 benchmark for gaming PCs, reportedly highlighted a 36 % efficiency enchancment. It was evaluated in Efficiency, Excessive, and Extremely situations. In the meantime, the DirectX 12 TimeSpy benchmark revealed a 33 % enchancment over the RTX 4090 GPU on Efficiency and Excessive settings.

Different reviewers additionally ran benchmark assessments reminiscent of SpeedWay, Port Royal, and Metal Nomad and got here to the conclusion that the general efficiency acquire could possibly be as excessive as 46 %. It ought to be famous that these are total graphics scores and never the chip’s total efficiency numbers. There is likely to be discrepancy between numbers achieved in artificial benchmarks and real-world gaming situations.

Nvidia GeForce RTX 5090 GPU Specs

Nvidia’s GeForce RTX 5090 GPU is outfitted with 32GB of GDDR7 VRAM (up from 24GB of GDDR6 VRAM in RTX 4090 GPU), has a peak clock pace of two.4GHz, and will get 21,760 CUDA cores. It’s constructed utilizing Nvidia’s Blackwell structure with fifth era Tensor cores and fourth era ray tracing cores.

General, the GPU options 92 billion transistors, able to delivering 3,352 trillion operations per second (TOPS). Nvidia claims that it’s the firm’s strongest GPU for client PCs thus far.