RTX 5070 vs RTX 3080
The RTX 5070 is about 19% faster than the RTX 3080. Here is where that gap comes from, what each one draws to get there, and whether it is the difference you are actually paying for.
RTX 5070 +19%
Measured against the RTX 3080 across the catalogue's combined workload index.
Where the gap is
Each row is scaled to the larger of the two values, so the bars show how much of the leader the other one reaches — not a shared scale across rows. The real figure is printed on every bar.
Side by side
| Specification | RTX 3080 | RTX 5070 |
|---|---|---|
| Performance index | 7400 | 8800 |
| Memory | 10 GB | 12 GB |
| Power draw | 320 W | 250 W |
| Performance per watt | 23.1 | 35.2 |
| Architecture | — | Blackwell |
| Generation | RTX 30 Series | RTX 50 Series |
How the gap is worked out
The percentage compares the two parts on the same relative performance index the site's bottleneck and build tools use, expressed as a share of the slower one. It is a single combined figure, not a frame rate: a specific game at a specific resolution can and will land either side of it, and a card with more memory pulls further ahead the moment a title needs that memory. Treat it as the size of the gap, not as a measurement of your machine.
What each one needs to run
The RTX 5070 and the RTX 3080 do not draw the same power, and the difference often outlives the card — a supply is the part you keep. Each page works the total out for a complete system.
Pair it with a processor
A faster card only shows up as faster frames if the processor can keep up with it. The bottleneck detector takes both and says which one is holding the other back.
Open the bottleneck detector