The NVIDIA GeForce RTX 4080 Super is finally here and at US$999 (~RM4,730) a pop for the card, we’re here to answer the obvious question of whether or not it’s worth pulling the US$999 (~RM4,725) trigger for. So, without further ado, and instead of wasting time here, I’m just going to dive right into it.
Specifications
Design
Then as now, as it is with all Founders Edition (FE) cards, this variant of the RTX 4080 Super comes with the same Dual Axis Flowthrough cooler shroud and cooling solution as its non-Super sibling. Like the RTX 4070 Super FE, the major change here is that the shroud that envelops the PCB is now all-black, rather than being a mix of gray with black fins.
Underneath the hood, the RTX 4080 Super FE can boast a slightly higher CUDA core count, base and boost clocks, shader multiprocessors (SMs), and four additional RT cores. That, and a slightly souped up 23Gbps frequency, an increased bandwidth of 736.3GB/s, and an L1 Data Cache exceeding one million kilobytes.
Beyond that, the RTX 4080 Super FE still has the same 16GB GDDR6X, as well as the same AD103 GPU architecture as the non-Super RTX 4080.
Testbench
In this review, I am testing the RTX 4080 Super FE and comparing its performance against the other RTX 40 Super Series GPUs and more importantly, against the non-Super RTX 4080. On that note, I have also retested the GPUs with the current GeForce drivers, given that it has been some time since the last time I tested it and improvements to such cards have clearly improved with the most up-to-date drivers. That driver version, by the way, is 551.22, which was provided to me by NVIDIA.
I have also streamlined the titles that I use for the gaming portion of the benchmarks and that is because, for the RTX 4080 Super and the rest of its RTX 40 Series siblings, they’ve basically reached a tipping point with regards to the frame rates, with or without enabling that pretty picture function that ray-tracing.
Benchmarks, Temperature, And Power Consumption
And the performance gains that the RTX 4080 Super has over its non-Super brethren is…not that far, it seems. In 3DMark, it’s really only pulling ahead by a very small margin and surprisingly, it’s an all-around gain.
In gaming, the RTX 4080 Super FE is consistent with its performance gains over the non-Super RTX 4080 in the majority of game, with the occasional “leaving the crowd in a blaze” with some titles, such as Starfield and, to my surprise, Cyberpunk 2077. That said, it also seems to stutter and lose its composure in the latter title against its sibling, and at 4K resolution too.
But again, for the most part, the RTX 4080 Super maintains a healthy framecount that sits above 100 fps for, once again, the majority of games, and well above the 60 fps base line that we seek out in all our games.
The RTX 4080 Super FE also seems to run considerably cooler than the non-Super RTX 4080. I am talking a whole 6°C cooler under load, which is actually mildly impressive.
Conclusion
At this stage, I’m starting to see why NVIDIA is doing what it is doing with the GeForce RTX 4080 Super FE and its non-Super sibling. Firstly, replacing the latter with the Super variant and essentially at same price when it launch is, in its own way, telling the world that it is offering some boost in performance, regardless of how small that upgrade actually is. The second point to this, though, is that by doing this, the brand is effectively cannibalising its own card and to that end, it raises the question as to why. That being said, this isn’t the first time we’ve seen this sort of behaviour from NVIDIA.
But let me be clear: if you’re trading up from an RTX 30 series GPU or older, and you are willing to pay the near RM5,000 price tag that NVIDIA’s AIB partners are going to charge for their custom version of the RTX 4080 Super, then yes, getting this card will send those framerates in your game soaring to new heights, and even sustain them. If you are thinking of trading up from that RTX 4070 Ti or 4080, though, then hear this: you’re good where you are.
