Nvidia has been slowly adding to its RTX 4000 Series of graphics cards, with the release of the RTX 4070 and RTX 4070 Ti.
Gamers and creatives alike may be wondering which type of GPU is best for them, and what kind of performance they can expect from the RTX 4070 compared to the RTX 4080.
We’re going to be running through the RTX 4070 and the RTX 4080 to see if more power is really worth it. Keep reading to find out all the differences between these GPUs, including any important similarities.
The RTX 4080 has more CUDA cores
The RTX 4080 comes with an impressive 9728 CUDA cores. In a nutshell, CUDA cores are Nvidia’s GPU equivalent of CPU cores, with more cores generally providing better performance. The RTX 4070 comes with 5888 cores, which will still be able to provide a lot of power.
Both of these cards support ray tracing, with a quoted 2x performance boost when compared to its predecessor. And DLSS 3 is also supported, using AI to significantly boost the frame rate of supported games.
The RTX 4070 has a lower boost clock speed
Boost clock speeds give a practical estimate of the ability of the GPU. This means that generally speaking, a graphics card with a higher boost clock speed should be more capable overall, although this is only really comparable for GPUs with the same architecture.
The RTX 4080 has a boost clock speed of 2.51GHz, while the RTX 4070 offers a boost clock speed of 2.48GHz. This will make the RTX 4080 the better choice for anyone wanting to engage in professional creative work and will be more capable of presenting AAA games at a high frame rate in 4K.
The RTX 4080 comes with more memory
Both graphics cards utilise the latest video memory standard in GDDR6X, but Nvidia has given the RTX 4080 slightly more memory.
The Nvidia RTX 4080 comes with a massive amount of vRAM at 16GB. While the RTX 4070 comes with 12GB RAM, which should still be enough for new titles.
What difference will this make? It’s hard to quantify, especially since both amounts are comfortably enough for modern games. But it’s still a point in the RTX 4080’s favour, once again showing it to be the more powerful GPU.
Both are built on Ada Lovelace architecture
These GPUs do share some similarities; both the RTX 4080 and RTX 4070 are built on the latest Nvidia Ada Lovelace architecture. This codename refers to the architecture featured on every GPU in the RTX 4000 Series, being built on a 4nm process node.
It comes with DLSS 3, which delivers up to 4x times the upscaling performance compared to its predecessor thanks to the 4th-gen Tensor Cores and Optical Flow Accelerator. 3rd-generation ray tracing cores have been added which should provide a more efficient performance when providing accurate and realistic shadows and lighting effects.
It’s important to note that while these GPUs do come with the same architecture, they will still perform differently. The inclusion of more CUDA cores and a higher boost clock speed on the RTX 4080 should make it the most powerful out of the two, even if the RTX 4070 still reaps the benefits of Lovelace.
