How the RTX 4090 and 4080 Evaluate to the RTX 3090 Ti and 3080 Ti
With the RTX 40-series set to launch lower than a month from now, let’s check out the specs, dig into the standout options, and see how the primary GPUs within the RTX 40 household stack up in opposition to the RTX 3090 Ti and the RTX 3080 Ti.
RTX 40 vs. RTX 3090 Ti and 3080 Ti: Temporary Overview
Because the infographic above notes, many modifications are coming from the RTX 30-series to the RTX 40-series. Nevertheless, the specs will not be all totally different from the Ti variants of the RTX 3090 and 3080, and the RTX 40-series all use GDDR6X reminiscence.
Nvidia claims that the RTX 4080 variants will present two to 4 occasions the efficiency of the RTX 3080 Ti. In the meantime, the RTX 4090 Ti is claimed to be two to 4 occasions quicker than the RTX 3090 Ti.
Different similarities price mentioning embrace the GPUs supporting hardware-based ray tracing along with Nvidia’s supersampling tech, DLSS. Whereas the RTX 30-series helps DLSS, there are some variations that we’ll talk about within the subsequent part.
One other fascinating notice is that there are two variants of the RTX 4080: one with 16GB of GDDR6X reminiscence, which can retail for $1,199, and one other with 12GB of GDDR6X reminiscence that may price $899. This may not be the primary time Nvidia did this, as the corporate launched two totally different variations of its RTX 3080 graphics playing cards – the launch mannequin included 10GB of reminiscence, whereas one other model launched practically two years later included 12GB of reminiscence.
However reminiscence (and pricing) isn’t the one distinction between the 2 RTX 4080 variants. The 16GB mannequin has extra RT, Tensor, and CUDA cores and requires at the least 750W of energy. In distinction, the 12GB mannequin has the next base and enhance clock and requires at the least 700W of energy.
RTX 40 vs. RTX 3090 Ti and 3080 Ti: Structure
The RTX 30-series is constructed on Nvidia’s Ampere graphical structure, which makes use of Samsung’s 8nm expertise. The RTX 40-series in the meantime is constructed on the brand new Ada Lovelace structure, which makes use of TSMC’s 4nm N4 expertise.
The Ampere structure is the second technology of RTX, which makes use of third-generation Tensor and second-generation RT cores. The Ada Lovelace structure, named after the mathematician and author, is the third technology of RTX, utilizing fourth-gen Tensor cores and third-gen RT cores. And as you count on with a brand new technology of microarchitecture, each Lovelace RT and Tensor core is extra highly effective than each Ampere RT and Tensor core.
Lovelace structure can also be the one graphical structure to help the brand new third technology of DLSS. Nvidia famous in a weblog publish that DLSS 3 is “powered by the brand new fourth-generation Tensor Cores and Optical Movement Accelerator,” that are discovered solely within the RTX 40-series.
In fact, these comparisons solely inform the story on paper, and we gained’t understand how the GPUs actually evaluate till we’ve the chance to benchmark the RTX 40-series ourselves. Nonetheless, the brand new fabrication course of and graphical structure present promising indicators that this can possible be a pleasant leap in efficiency between the 2 generations of GeForce graphics playing cards.
Taylor is the Affiliate Tech Editor at IGN. You’ll be able to observe her on Twitter @TayNixster.