Introduction
At CES 2024, NVIDIA announced its SUPER series refresh for the GeForce RTX 40 series GPUs. The new SUPER series is comprised of the GeForce RTX 4070 SUPER, GeForce RTX 4070 Ti SUPER, and the GeForce RTX 4080 SUPER. We’ve recently reviewed the GeForce RTX 4070 SUPER and GeForce RTX 4070 Ti SUPER, and today we will review the GeForce RTX 4080 SUPER and put it up against the GeForce RTX 4080 and AMD Radeon RX 7900 XTX in 10 games, testing Upscaling and Ray Tracing at 4K.
Back in January, we looked at an overview of the GeForce RTX 4080 SUPER specifications, and now we will take a closer look at what separates it in the lineup. The most notable specification is actually the pricing. NVIDIA has set the MSRP for the GeForce RTX 4080 SUPER at $999, a 17% reduction in pricing from the $1,199 price tag of the GeForce RTX 4080. This alone, is more appealing, but NVIDIA didn’t quite stop there, this isn’t exactly a re-brand of the GeForce RTX 4080, the GeForce RTX 4080 is actually different in some regards, but the question is will it really make a difference in gameplay performance? We shall see.
NVIDIA GeForce RTX 4080 SUPER Specs
The GeForce RTX 4080 SUPER is based on the full die of AD103, whereas the GeForce RTX 4080 was slightly cut down from the full die specs. The GeForce RTX 4080 SUPER, therefore, provides the full SM count and more CUDA cores and Tensor Cores. The memory frequency is also increased. All of this is of course based on the Ada Lovelace architecture and all the features within. What’s even more fascinating, is that the TGP remains the same at 320W just like the GeForce RTX 4080, meaning it is very efficient, and perhaps binned better.
The GeForce RTX 4080 SUPER has 80 SMs and 10,240 CUDA Cores which gives it 80 RT Cores and 320 Tensor Cores, 320 Texture Units, and 112 ROPs. The base clock is 2295MHz while the Boost Clock is 2550MHz. It has 65536K of L2 cache. It has 16GB of GDDDR6 at 23Gbps on a 256-bit memory bus giving it 736GB/s of memory bandwidth. The TGP is 320W.
| Specifications | GeForce RTX 4080 | GeForce RTX 4080 SUPER |
|---|---|---|
| Die | AD103 | AD103 |
| SMs | 76 | 80 |
| CUDA Cores | 9728 | 10240 |
| RT Cores | 76 | 80 |
| Tensor Cores | 304 | 320 |
| ROPs | 112 | 112 |
| TMUs | 304 | 320 |
| Base Clock | 2205MHz | 2295MHz |
| Boost Clock | 2505MHz | 2550MHz |
| Memory | 16GB GDDR6X | 16GB GDDR6X |
| Bus Width | 256-bit | 256-bit |
| Memory Clock | 22.4Gbps | 23Gbps |
| Memory Bandwidth | 716GB/s | 736GB/s |
| L2 Cache | 64MB | 64MB |
| TGP | 320W | 320W |
From the table above, you can see the specification differences between the GeForce RTX 4080 and GeForce RTX 4080 SUPER. With 4 more SMs, the GeForce RTX 4080 SUPER has a small 5% increase in CUDA Cores, a 5% increase in Tensor Cores, and a very small 2% boost in clock speed as well as a 3% increase in memory clock and 3% increase in memory bandwidth. 
While yes, the GeForce RTX 4080 SUPER has higher specs than the GeForce RTX 4080, on paper, these are very small differences, and so we will have to see how that translates to performance, which may be very minor. At any rate, one great thing is that the power has not increased through all this, and so the demand on power does not increase, all while the price does decrease, making it more appealing.
NVIDIA GeForce RTX 4080 SUPER Founders Edition Pictures
The NVIDIA GeForce RTX 4080 SUPER Founders Edition is similar in size to the GeForce RTX 4080. It measures 12.2″ inches in length, 5.5″ inches in width, and 2.4″ in height, so it is a thick video card. It does, however, have an aesthetic change, with an all-black-out look to it, which is appealing visually.
NVIDIA uses the 12VHPWR connector, and an adapter cable with 3x 8pin PCIe power connectors is included. For display out, you’ll find 1x HDMI and 3x DisplayPort. The fans operate in a push-pull configuration as we are accustomed to on this design. It’s quite eloquent and sturdy.
When installed in the computer, you will find an LED logo atop the video card, as well as an LED feature on the front of the video card.