NVIDIA GeForce RTX 4090 vs NVIDIA GeForce RTX 5080
Side-by-side benchmark comparison, specs, pricing, and value analysis.
Benchmark Comparison
Gaming figures are average frames per second, measured by Tom's Hardware GPU hierarchy. Geometric mean of 15 games, ultra settings, no upscaling. Ray tracing figures are DXR-enabled at 1440p. A dash means the card was not tested at that resolution. Productivity and Steel Nomad are synthetic scores on their own scales and are not comparable to the fps columns.
Specifications
| Spec | GeForce RTX 4090 | GeForce RTX 5080 |
|---|---|---|
| VRAM | 24GB GDDR6X | 16GB GDDR7 |
| Memory Bandwidth | 1008 GB/s | 960 GB/s |
| Shaders | 16384 | 10752 |
| Boost Clock | 2520 MHz | 2617 MHz |
| TDP | 450W | 360W |
| Architecture | Ada Lovelace | Blackwell 2.0 |
| 3DMark Steel Nomad | 9,236 | 8,762 |
| Best Price | $1,599 | $999 |
| Value Score | 27/100 | 46/100 |
Verdict
The NVIDIA GeForce RTX 4090 is faster but costs more. The NVIDIA GeForce RTX 5080 offers better value for money. Your choice depends on budget vs. performance priority.
NVIDIA GeForce RTX 4090 vs NVIDIA GeForce RTX 5080: Which Should You Buy?
The NVIDIA GeForce RTX 4090 is 6% faster at 1440p ultra (146.1 fps against 138.4 fps). That is roughly one settings notch — enough to matter if you are chasing a specific refresh rate, not enough to rescue a bad price.
You pay a 60% premium for the NVIDIA GeForce RTX 4090 ($1,599 against $999) to get 6% more frames. The extra frames do not keep up with the extra cost — the NVIDIA GeForce RTX 5080 is the better value here.
Two things beyond raw speed: NVIDIA GeForce RTX 4090 carries 24GB against 16GB — the deciding factor at 1440p and above with ray tracing or texture packs, and the reason 16GB cards age badly; and the NVIDIA GeForce RTX 5080 draws 90W less (360W versus 450W), which can mean a smaller PSU and a quieter case; and the NVIDIA GeForce RTX 5080 is 3 years newer (2025 versus 2022), so it gets longer driver support and newer upscaling and frame generation features.