Early benchmarks of the highly anticipated RTX 5090 have revealed some surprising findings that contrast with NVIDIA’s marketing claims. The performance gains over its predecessor are not quite what was initially promised.
Real-World Performance
The RTX 5090 demonstrates approximately 30% better performance compared to the RTX 4090, which was released over two years ago. This improvement comes at the cost of significantly higher power consumption, with the GPU drawing an average power of 492W during testing.
Marketing vs. Reality
NVIDIA’s CEO Jensen Huang’s claim of the RTX 5090 being twice as fast as the RTX 4090 appears to be misleading. The only scenario where the GPU approaches this claimed 2x performance boost is when using DLSS Frame Generation technology.
DLSS 4 Challenges
While DLSS technology continues to evolve, the current implementation shows several visual artifacts:
- Visible combing effects on crosswalks in Cyberpunk
- Bottles phasing in and out during benchmarks
- Doubled fan blades in certain scenarios
- Noticeable artifacting around player models and flashlight edges
Technical Analysis
The performance gains are particularly underwhelming considering both the RTX 5090 and RTX 4090 utilize the same manufacturing node. This raises questions about the architectural improvements and efficiency gains in the new generation.
Conclusion
While the RTX 5090 does offer performance improvements over its predecessor, the actual gains fall significantly short of NVIDIA’s marketing claims. The combination of higher power consumption and visual artifacts with DLSS Frame Generation suggests that potential buyers should carefully consider their upgrade decisions based on real-world benchmarks rather than marketing promises.
