After months of rumors and teasing, Nvidia finally announced the RTX 40-series of graphics cards during its GTC 2022 keynote. Using the caller Ada Lovelace graphical architecture, the RTX 40-series volition service arsenic a successor to the RTX 30-series, which debuted backmost successful 2020.

With the RTX 40-series acceptable to motorboat little than a period from now, let’s instrumentality a look astatine the specs, excavation into the standout features, and spot however the archetypal GPUs successful the RTX 40 household stack up against the RTX 3090 Ti and the RTX 3080 Ti.

RTX 40 vs. RTX 3090 Ti and 3080 Ti: Brief Overview

As the infographic supra notes, galore changes are coming from the RTX 30-series to the RTX 40-series. However, the specs are not each antithetic from the Ti variants of the RTX 3090 and 3080, and the RTX 40-series each usage GDDR6X memory.

Nvidia claims that the RTX 4080 variants volition supply 2 to 4 times the show of the RTX 3080 Ti. Meanwhile, the RTX 4090 Ti is claimed to beryllium 2 to 4 times faster than the RTX 3090 Ti.

Other similarities worthy pointing retired see the GPUs supporting hardware-based ray tracing successful summation to Nvidia’s supersampling tech, DLSS. While the RTX 30-series supports DLSS, determination are immoderate differences that we’ll sermon successful the adjacent section.

Another absorbing enactment is that determination are 2 variants of the RTX 4080: 1 with 16GB of GDDR6X memory, which volition retail for $1,199, and different with 12GB of GDDR6X representation that volition outgo $899. This would not beryllium the archetypal clip Nvidia did this, arsenic the institution released 2 antithetic versions of its RTX 3080 graphics cards – the motorboat exemplary included 10GB of memory, portion different mentation released astir 2 years aboriginal included 12GB of memory.

But representation (and pricing) isn’t the lone quality betwixt the 2 RTX 4080 variants. The 16GB exemplary has much RT, Tensor, and CUDA cores and requires astatine slightest 750W of power. In contrast, the 12GB exemplary has a higher basal and boost timepiece and requires astatine slightest 700W of power.

RTX 40 vs. RTX 3090 Ti and 3080 Ti: Architecture

The RTX 30-series is built connected Nvidia’s Ampere graphical architecture, which uses Samsung’s 8nm technology. The RTX 40-series meantime is built connected the caller Ada Lovelace architecture, which uses TSMC’s 4nm N4 technology.

The Ampere architecture is the 2nd procreation of RTX, which uses third-generation Tensor and second-generation RT cores. The Ada Lovelace architecture, named aft the mathematician and writer, is the 3rd procreation of RTX, utilizing fourth-gen Tensor cores and third-gen RT cores. And arsenic you expect with a caller procreation of microarchitecture, each Lovelace RT and Tensor halfway is much almighty than each Ampere RT and Tensor core.

Lovelace architecture is besides the lone graphical architecture to enactment the caller 3rd procreation of DLSS. Nvidia noted successful a blog station that DLSS 3 is “powered by the caller fourth-generation Tensor Cores and Optical Flow Accelerator,” which are recovered exclusively successful the RTX 40-series.

Of course, these comparisons lone archer the communicative connected paper, and we won’t cognize however the GPUs genuinely comparison until we person the accidental to benchmark the RTX 40-series ourselves. Still, the caller fabrication process and graphical architecture amusement promising signs that this volition apt beryllium a bully leap successful show betwixt the 2 generations of GeForce graphics cards.

Taylor is the Associate Tech Editor astatine IGN. You tin travel her connected Twitter @TayNixster.