All the details about the new gaming GPU range

The age of Ada is drawing to a close, and the new Nvidia GeForce RTX 5000 range is going to be unleashed very soon, but what do Jensen and his team of GPU architects have in store for PC gamers this time? Nvidia lifted the lid on four new gaming GPUs based on the Blackwell architecture at CES, and we’ve summarized all the key details below.

Now we know that AMD is abandoning high-end GPUs for the next Radeon generation, Nvidia looks set to have a clean run at the premium end of the GPU market. Assuming there are no catastrophic disasters on the way, this means Nvidia will be able to grab the best graphics card title with its next flagship GPU, the RTX 5090.

Nvidia GeForce RTX 5000 GPU range

The initial wave of Nvidia GeForce RTX 5000 graphics cards comprises the RTX 5090 and 5080 at the high end, with the RTX 5070 Ti and 5070 occupying the mid-range sector.

Nvidia GeForce RTX 5090

The cream of the crop in Nvidia’s new Blackwell lineup is the RTX 5090. This new GPU comes with a massive 32GB of GDDR7 VRAM, along with 21,760 CUDA cores, which already makes it more powerful than the mighty RTX 4090.

However, Nvidia also has another trick up its sleeve with DLSS 4, which supports Multi Frame Gen to boost frame rates further. We had a play with this in Cyberpunk 2077, Marvel Rivals, and Indiana Jones and the Great Circle at CES, and found that it worked remarkably well. In games that support this feature, Nvidia says you’ll be able to play games at 4K with a frame rate of 240fps.

The RTX 5090 VRAM is also attached to a super-wide 512-bit memory interface, meaning this GPU has an enormous amount of memory bandwidth at its disposal, 1,792GB/s to be precide. The RTX 5090 power draw is also high at 575W, meaning you’ll need at least a 1,000W PSU to run this card, although it still only requires a single 12VHPWR connector.

For more details about the specs, price, and release date, check out our full RTX 5090 guide.

Nvidia GeForce RTX 5080

The second Blackwell GPU from the front is the new RTX 5080, which has a significantly lesser spec than the 5090. The specs list shows that the RTX 5080 has 10,752 CUDA cores at its disposal, which is more than the RTX 4080, but half the number found in the 5090.

You also only get half the memory of the 5090, at 16GB, but the use of high-speed 32Gbps GDDR7 VRAM means the memory bandwidth is much improved. Even with just a 256-bit memory interface,the RTX 4080 has a total memory bandwidth of 960GB/s, which isn’t far off the RTX 4090. In its favor, the RTX 5080 has a significantly lower power draw of 360W.

There are also rumors that there will be a 24GB RTX 5080 coming at a later date, which will reportedly have the same core GPU spec as the 16GB card, just with more memory. This would be possible using 3GB rather than 2GB DRAM chips, such as the new Samsung 24Gb GDDR7 DRAM that’s recently been announced.

For more details about the rumored specs, price, and release, read our full RTX 5080 guide.

Nvidia GeForce RTX 5070

Nvidia claims the RTX 5070 is just as fast as the RTX 4090, and while this claim deserves a healthy dose of skepticism, we struggled to tell the difference between the two cards when running Marvel Rivals in our RTX 5070 vs 4090 test.

The key behind this trick is DLSS 4 Multi Frame Gen support, which enables the RTX 5070 to run at much higher frame rates than the RTX 4070, despite only having a little more power under the hood.

The RTX 5070 has 6,144 CUDA cores at its disposal, and its use of GDDR7 VRAM also means it has a decent 672GB/s of memory bandwidth, despite its comparatively narrow 192-bit bus. However, it’s also rumored sadly only equipped with 12GB of VRAM, much like the RTX 4070 today.

For more details about the rumored specs, price, and release, you can read our full RTX 5070 guide.

Nvidia GeForce RTX 5060

The rumored RTX 5060 is set to be the most affordable desktop GPU in the new Blackwell lineup, and it’s expected to be launched shortly after the RTX 5070. There haven’t been many leaks about the specs yet, but we hope that this GPU will fix some of the shortcomings of its predecessor, the RTX 4060, such as its lack of bandwidth and surprisingly high price.

For more details about the rumored specs, price, and release, check out our full RTX 5060 guide.

Nvidia GeForce RTX 5000 price

The RTX 5000 price currently starts at to start at $549 for the RTX 5070 and goes up to $1,999 for the 5090. 

While you’re waiting for the new GeForce RTX 5000 lineup to arrive, check out our GeForce RTX 4080 Super review, where we put Nvidia’s current $999 champion through its paces.