Nvidia GeForce RTX 5070 review: DLSS 4 doesn’t deliver 4090 performance

Despite the overwhelming power of the RTX 5090, it was perhaps the RTX 5070 that saw the most headlines in the wake of Nvidia’s 50-series announcements. RTX 4090 levels of performance from a $550 card is a heck of a promise, but what’s arrived isn’t anywhere near as exciting. Instead, we’re looking at something akin to the RTX 4070 Super, a good card by all accounts at its $600 price point, but equipped with DLSS 4 and discounted to $550. That’s a reasonable upgrade, but not one that’ll grab the headlines.

After all, as we’ve spent more time with DLSS 4, it’s become clear that lower-end tiers see less utility from the multi frame generation feature, requiring more setup to keep latency at acceptable levels. It’s still a great option to have – and we get into how it operates in more detail on page eight – but it’s not quite the set-and-forget ticket for maxing out your monitor’s refresh rate as it is on the 5080 and 5090.

There are other by-now familiar limitations too, with the decent gen-on-gen gains of the RTX 5090 getting less and less impressive the lower we go on the stack. Fundamentally, this is a relatively restrained upgrade, built on the same process as 40-series, with the gen-on-gen upgrades coming just from faster GDDR7 memory and higher power budgets. So too is it with the 5070, with performance regressions in a handful of games versus 4070 Super, along with six to seven percent performance wins in most other titles.

Nvidia faces other problems too – beyond the reports of black screen issues, missing ROPs and the like. AMD has announced two GPUs: 9070 and 9070 XT. Extrapolating out from AMD’s benchmarks, the 9070 has more memory and more performance for the same $550 sticker price – yes, even ray tracing. Meanwhile, pay a $50 premium and you get an AMD GPU that is basically as potent as a 5070 Ti.

If you’re reading this review on day one, you’ll probably know that the 9070 and 9070 XT are still embargoed, but expect these graphics cards to play a heavy part in the 5070 value discussions over the days, weeks and months ahead.

One thing that isn’t up for much debate is the quality of the RTX 5070’s Founders Edition design, which returns after being absent for the RTX 5070 Ti. The two-slot form factor looks great, is delivered in clever sustainable packaging and performs with the minimum of noise – though overclocking does raise fan speeds to noticeable levels.

In terms of overclocks, a higher 110 percent power budget, 350MHz core overclock and 500MHz memory overclock delivered around seven percent higher performance for seven percent more power – reasonable stuff, but the 5070 Ti is still faster and more efficient thanks to its larger GB203 processor.

You can see more of of our overclocked results in the embedded video above, but for now we’ll return to the stock clocks for the remainder of our testing. Alongside it, we’re using a top-end system based around the fastest gaming CPU, the AMD Ryzen 7 9800X3D, to shift the burden to the graphics card as much as possible. We also have 32GB of Corsair DDR5-6000 CL30 memory, a high-end Asus ROG Crosshair X870E Hero motherboard and a 1000W Corsair PSU.

With all that said, let’s get into the benchmarks.

Nvidia GeForce RTX 5070 Analysis

• Introduction [This Page]
• RT benchmarks: Alan Wake 2, Avatar: Frontiers of Pandora, Cyberpunk 2077
• RT benchmarks: Dying Light 2, F1 24, Hitman: World of Assassination
• RT benchmarks: Indiana Jones and the Great Circle, Metro Exodus Enhanced Edition, A Plague Tale: Requiem
• Game benchmarks: Alan Wake 2, Black Myth: Wukong, Cyberpunk 2077
• Game benchmarks: F1 24, Forza Horizon 5, Senua’s Saga: Hellblade 2
• Game benchmarks: Hitman: World of Assassination, A Plague Tale: Requiem
• DLSS 4 and Path Tracing: Cyberpunk 2077, Alan Wake 2
• Conclusions, value and recommendations