Nvidia GeForce RTX 5090 Review

The RTX 5090 is costly but offers moderate improvement over the RTX 4090.

: Nvidia's GeForce RTX 5090, priced at $2,000, is the world's fastest gaming GPU but offers only a 27% performance increase over its predecessor, the RTX 4090. It uses the same TSMC 4N process as before and includes 32 GB of 28Gbps GDDR7 memory on a 512-bit bus. While its thermal efficiency is commendable, its value is questionable given the steep price. The review suggests it feels more like a '4090 Ti' due to underwhelming performance gains.

Nvidia's latest GeForce RTX 5090, tagged at $2,000, emerges as the fastest gaming GPU on the market, yet its performance uplift over the RTX 4090 stands at an average of 27%. This increase is coupled with a hefty price which is 25% more than its predecessor, raising questions about its value in the eyes of consumers.

Despite leveraging cutting-edge components like 32 GB of 28Gbps GDDR7 memory and offering a notable 33% increase in cores, the card still utilizes the previous TSMC 4N process without node improvements. Nvidia's challenge this generation was substantial; to deliver marked enhancements without architectural overhauls, leading to criticisms that it resembles a model variant of the RTX 4090 rather than a new generation flagship.

The RTX 5090's performance tests highlighted its strengths in thermal management and power efficiency, maintaining functionality even under higher thermal loads. Many observed benchmarks, however, showcased only a modest improvement over the RTX 4090, suggesting its flagship status could be more marketing than genuinely revolutionary technology. Ray tracing performance improvements were similarly modest, further casting doubt on its proposition as a substantial leap forward.