NVIDIA has published on its official website some new benchmarks linked to the performance improvements of its technology DLSS 3. Thanks to this, you can already have a pretty good idea of the difference in performance between the GeForce RTX 4080 16GB and GeForce RTX 4080 12GB.
Remember that except for the name of the GPU, There is little in common between these two GeForce RTX 4080, and is that both use quite different silicon. Where the 16 GB capacity variant boasts of offering 9,728 CUDA Coreswhile the 12 GB variant sees its computing power significantly reduced by having 7,680 CUDA Cores. What is summarized in this difference of nuclei? Well now we will see it in great detail.
The NVIDIA GeForce RTX 4080 16GB is up to 30% faster than its little brother
According to the company’s own evidence, the performance difference between both GPUs is quite remarkable despite the fact that both they share the same name model, which is very misleading.
With a Intel Core i9-12900K32 GB of RAM and Windows 11, the GeForce RTX 4080 16 GB offers a performance of 66.4fps in Microsoft Flight Simulator at a 4K resolution of 3840 x 2160 pixels, which scales up to 147.3FPS with DLSS 3. The GeForce RTX 4080 12 GB offers performance of 52.7/121.5fps.
In A Plague Tale: Requiem, the 16 GB variant offers a performance of 55.2/128.6FPS compared to a 12 GB variant that offers 42.4/106.3FPSwhile in the F1 22 the performance difference is 57/143fps versus 46/113fps.
In essence, adding all these tests, we talk about the GeForce RTX 4080 16 GB being 26% faster in Microsoft Flight Simulator, 24% faster in F1 22, and 30% faster in A Plague Tale: Requiem. If we put the RTX 4090 in the equation, it becomes up to 151% faster than the 12GB RTX 4080.
GeForce RTX 4080 16GB vs GeForce RTX 4080 12GB
The GeForce RTX 4080 with 16GB of GDDR6X memory, in addition to offering greater capacity, it also reaches a higher speed of 22.5Gbps. This is attached to a 256-bit interface, which gives us a bandwidth of 720GB/s. This memory is attached to a graphics chip that offers 9,728 CUDA Cores at a Base/Turbo frequency of 2.21/2.51GHz with a TDP is 320Wbeing able to be squeezed until reaching a maximum consumption of 516W.
On the other hand, we have a GeForce RTX 4080 with 12GB GDDR6X memory at 21Gbps. In addition to having less memory, and being slower, it is also equipped with a memory interface of 192 bitwhich reduces your bandwidth to 504GB/s. This model has 7,680 CUDA Cores at a Base/Turbo frequency of 2.31/2.61GHzand manages a TDP of 285W (366W maximum consumption).
Both GPUs will go on sale in November at a suggested starting price of €1,099 for the RTX 4080 12 GB and from €1,499 for the RTX 4080 16GB. In this case, we are talking about being up to 30% faster at a cost of 36% more. For reference, you can currently find the GeForce RTX 3090 Ti by 1,300 euros with 24 GB of memory and positioning itself between the RTX 4080 12 GB and 16 GB.