The Nvidia GeForce RTX 5070 is now officially out in the wild, if you’re lucky enough to find one in stock. Nvidia took the wraps off its latest mid-range gaming GPU at the CES 2025 trade show on January 6 and, despite it having specs that largely align with the RTX 4070, Nvidia claimed this new GPU could offer the performance of an RTX 4090.
However, in our Nvidia GeForce RTX 5070 review, we found this claim didn’t hold up to much scrutiny. Yes, the RTX 5070 can ostensibly beat the RTX 4090 in games such as Cyberpunk 2077, but it can only do it by using multi frame gen, a new tech from Nvidia that uses AI to insert up to three AI-generated frames between each pair of genuinely rendered frames.
Launching alongside the colossally powerful new flagship RTX 5090 – a GPU that now sits atop our best graphics card guide thanks to its stunning performance in our RTX 5090 review – and the RTX 5080, this new mid-range card is priced at just $549. However, it’s up against fierce competition from AMD now, as you can see in our Radeon RX 9070 XT review.
Nvidia GeForce RTX 5070 release date
The RTX 5070 release date was Wednesday, March 5, 2025, just a month after the release of the RTX 5090, and far sooner than is typical for a mid-range Nvidia GPU release.
Nvidia GeForce RTX 5070 price
The RTX 5070 price is $549, which is slightly less than many had predicted, and $50 less than the price of the RTX 4070 at launch. However, there is tough competition in this price sector now, with AMD aiming its new Radeon RX 9070-series GPUs firmly in the RTX 5070’s face. The Radeon RX 9070 has the same $549 price as the RTX 5070, and the 9070 XT price is a little more expensive at $599.
Nvidia GeForce RTX 5070 specs
The Nvidia GeForce RTX 5070 specs include 6,144 CUDA cores and 50 RT cores, but it only has 12GB of GDDR7 VRAM.
| Nvidia GeForce RTX 5070 | |
| CUDA cores | 6,144 |
| Tensor cores | 192 (5th-gen) |
| RT cores | 48 (4th-gen) |
| Boost clock | 2.51GHz |
| Base clock | 2.32GHz |
| Interface | 16x PCIe 5.0 |
| VRAM | 12GB 28Gbps GDDR7 |
| Memory interface | 192-bit |
| Memory bandwidth | 672GB/s |
| Power draw | 250W |
| Power connector | 1 x 16-pin or 2 x 8-pin |
On paper, the Nvidia GeForce RTX 5070 specs look relatively modest, with it only housing a few hundred more CUDA cores than the RTX 4070 and having the same 192-bit memory interface and 12GB of VRAM. However, with the card using much faster GDRR7 memory, ita bandwidth is much higher.
Moreover, Nvidia has made the bold claim that the new card will offer “4090 performance for $549,” suggesting this new card will be as fast as the company’s old $1,599 flagship card. However, the big caveat here is that this refers to performance when using the latest DLSS 4 upscaling and multi-frame generation technology with neural rendering.
This enables up to three frames to be generated entirely by AI, allowing for incredible boosts to the perceived frame rate. We tested this RTX 4090 vs RTX 5070 scenario and DLSS 4 does indeed allow the RTX 5070 to offer a higher perceived frame rate in games that support multi-frame generation, but the 4090 is still far faster in most scenarios.
AMD is also talking up its new FSR 4 technology as a competitor to DLSS, which looks and performs well in our tests. The RX 9070 XT also has the necessary raw performance to outclass the RTX 5070 in titles that don’t support DLSS, making AMD’s card the more compelling option overall.
If you’re set on buying an RTX 5070, check out our where to buy 5070 guide that provides a daily update on the stock situation of the card. There’s not a lot of stock right now, but it’s worth keeping an eye on that page to see if the situation changes.
