NVIDIA unveils three new GeForce RTX 40 ‘Super Series’ GPUs

NVIDIA has unveiled the GeForce RTX 40 Super Series: the three latest graphics cards in its GeForce RTX 40 Series of consumer GPUs based on the Ada Lovelace GPU architecture.

The new cards – the GeForce RTX 4070 Super, GeForce RTX 4070 Ti Super and GeForce RTX 4080 Super – all offer performance improvements over their existing non-Super counterparts, the GeForce RTX 4070, GeForce RTX 4070 Ti and GeForce RTX 4080.

All three will be available later this month, with the existing GeForce RTX 4070 Ti and GeForce 4080 being phased out in favor of their new Super counterparts.

Although primarily gaming cards, NVIDIA now markets the RTX 40 Series at creators as well as gamers, so below, we’ve rounded up their key specifications for DCC work, and benchmark scores for CG software.

*Data taken from third-party websites.

Key specifications for DCC work
As you might expect, the three new GPUs have similar basic specifications to their non-Super counterparts, but are slightly to signficantly better.

For raw power, CUDA core counts – and corresponding FP32 compute power – Tensor core counts and RT core counts are all roughly 5-20% higher than the corresponding older cards.

The difference is most significant between the GeForce RTX 4070 Super and GeForce RTX 4070.

With the GeForce RTX 4070 Ti Super, another key advance over the existing GeForce RTX 4070 Ti is GPU memory, which increases from 12 GB to 16 GB, significantly increasing the maximum size of scenes that will fit in on-board memory during GPU rendering.

Memory bandwidth also increases: the GeForce RTX 4070 Ti Super has a 256-bit memory bus, compared to the 192-bit bus of the older card.

And with the GeForce RTX 4080 Super, a key selling point over the older GeForce RTX 4080 is the launch price: at $999, down $200 from that of its older counterpart.

Benchmarks and performance in DCC applications
Although most of the benchmark scores on NVIDIA’s product website are for games, they include a couple of applications more relevant to CG artists: Blender and image upscaling software On1 Resize AI.

The charts compare the new cards with previous-generation GeForce RTX 30 Series and 20 Series GPUs rather than their non-Super current-generation counterparts.

Not suprisingly, the new GPUs are faster than cards three to five years old, but in Blender, the biggest relative increase in performance is for the GeForce RTX 4070 Ti Super, which is shown as having performance just under 2x that of the previous-generation GeForce RTX 3070 Ti.

There’s no information about how the cards were tested on the product website, so we aren’t sure whether that reflects GPU rendering performance or viewport interactivity.

Updated 29 January 2024: NVIDIA tells us that the cards were tested with the official Blender benchmark, which measures path tracing render performance in Blender’s Cycles renderer.

Pricing and release dates
The GeForce RTX 4070 Super, GeForce RTX 4070 Ti Super and GeForce RTX 4080 Super will all be available later this month, with recommended prices of $599, $799 and $999, respectively.

Read NVIDIA’s announcement of the GeForce RTX 40 Super Series GPUs

Find specifications for the GeForce RTX 40 Super Series GPUs on NVIDIA’s website

Have your say on this story by following CG Channel on Facebook, Instagram and X (formerly Twitter). As well as being able to comment on stories, followers of our social media accounts can see videos we don’t post on the site itself, including making-ofs for the latest VFX movies, animations, games cinematics and motion graphics projects.