NVIDIA’s third and perhaps the final RTX 40 Super card for this generation has now been unleashed. We recently reviewed one in the form of MSI’s GeForce RTX 4080 Super Expert graphics card. In this article, let’s compare the RTX 4080 Super to the outgoing RTX 4080. What’s the difference between the RTX 4080 Super and the non-Super GPU? How much faster is the RTX 4080 Super vs RTX 4080? And is the RTX 4080 Super a better GPU? Well, spoiler alert, not much, actually. Both the RTX 4080 Super and RTX 4080 perform similarly. But if you want to see the breakdown or more details, please continue reading below.
NVIDIA GeForce RTX 4080 Super vs RTX 4080 Specs Comparison
Despite the “SUPER” branding, the new GeForce RTX 4080 Super isn’t that different from its predecessor, the RTX 4080. For starters, both use the same AD103 GPU, but the RTX 4080 Super has a slight bump in its specs. It has 10240 CUDA cores, a mere 5% increase from the RTX 4080’s 9728 cores. The RTX 4080 Super also has 320 Tensor cores, 80 RT cores, and a measly 45MHz increase in its boost clock speed.
Unfortunately, NVIDIA didn’t increase its VRAM. The RTX 4080 Super still has 16GB of GDDR6X memory but with a slightly higher clock speed and bandwidth.
The real difference between the RTX 4080 Super and the RTX 4080 is the starting price. The RTX 4080 launched with an MSRP of $1,199. Meanwhile, the RTX 4080 Super starts at $999. Many believed that the RTX 4080 should have launched at $999 in the first place.
Below are the specs of the RTX 4080 Super and RTX 4080. After that, let’s proceed with the benchmarks.
Graphics Card | RTX 4080 Super | RTX 4080 |
---|---|---|
Fabrication Process | TSMC 4 nm NVIDIA Custom Process | TSMC 4N NVIDIA Custom Process |
Transistor Count | 45.9 billion | 45.9 billion |
Architecture | Ada Lovelace | Ada Lovelace |
GPU | AD103 | AD103 |
Graphics Processing Clusters | 7 | 7 |
Texture Processing Clusters | 40 | 38 |
Streaming Multiprocessors | 80 | 76 |
CUDA Cores | 10240 | 9728 |
Tensor Cores | 320 (4th Gen) | 304 (4th Gen) |
RT Cores | 80 (3rd Gen) | 76 (3rd Gen) |
Texture Units | 320 | 304 |
ROPs | 112 | 112 |
Boost Clock | 2550 MHz | 2505 MHz |
Memory Clock | 11500 MHz | 11200 MHz |
Memory Data Rate | 23 Gbps | 22.4 Gbps |
L2 Cache Size | 65536 KB | 65,536 KB |
Total Video Memory | 16 GB GDDR6X | 16 GB GDDR6X |
Memory Interface | 256-bit | 256-bit |
Total Memory Bandwidth | 736 GB/s | 716.8 GB/s |
Texture Rate | 816 Gigatexels/sec | 761.5 Gigatexels/sec |
Connectors | 3x DisplayPort, 1x HDMI | 3x DisplayPort, 1x HDMI |
Minimum Power Supply | 750 Watts | 750 Watts |
Total Graphics Power (TGP) | 320 Watts | 320 Watts |
PCI Express Interface | Gen 4 | Gen 4 |
Release Date | January 2024 | November 2022 |
Launch Price | $999 | $1,199 |
GeForce RTX 4080 Super vs RTX 4080 Benchmarks
RTX 4080 Super vs RTX 4080 Synthetic Benchmarks
In the 3DMark benchmark suite, the RTX 4080 Super is just slightly faster than the RTX 4080. If this were a blind test, I’d say we are comparing the same GPU. Except that one GPU is an overclocked version, and the other is a stock model.
Next, let’s see if there’s any difference in actual games. Perhaps those extra cores may prove helpful in actual games.
RTX 4080 Super vs RTX 4080 Game Benchmarks
I only tested these GPUs in QHD 1440p and UHD 2160p. These GPUs are not meant for 1080p gaming. At 1080p, both 4080s will only be limited by the resolution, and we’ll only get a less accurate result. Both GPUs will have a better chance of showing their capabilities at higher resolution. Below are the results I got:
Unfortunately, even with actual games, there is no substantial difference between the RTX 4080 Super vs RTX 4080. Both GPU performs similarly. There’s even an instance where the RTX 4080 Super slightly fell behind the RTX 4080, which is unacceptable given its slightly “better” specs.
Summary Results
In general, the RTX 4080 Super is just around 1% to 2% faster than the RTX 4080. At best, and in some cherry-picked results, the RTX 4080 Super could be up to 3% to 5% faster. Still, those are not substantial performance improvements. Again, it looks like we are just comparing two graphics cards of the same GPU. Only that the slightly faster one is an overclock variant and the other is a reference model.
We could get that tiny improvement by manually overclocking the GPU or by fine-tuning the system. Not to mention, 1% to 2% usually falls within the margin of error, and usually, one benchmark run is slightly different from the second run.
Pricing Difference
So, we have determined that the RTX 4080 Super performs similarly to the RTX 4080. How about the price? Luckily, NVIDIA released the RTX 4080 Super at a lower price. The RTX 4080 Super starts at $999. But there are only a few MSRP cards available. I’m sure the Founders Edition will be out of stock immediately as scalpers tend to buy as many FE models as possible and resell them at a higher price.
Most RTX 4080 Super graphics cards from NVIDIA’s AIC partners are OC models or factory-overclocked cards. This means these graphics cards will be slightly higher than the starting price. However, premium models, like MSI’s SUPRIM X or Asus’ ROG Strix, are expected to be priced much higher. Expect to pay a premium for these higher-end models.
The RTX 4080 Super is about ~17% cheaper than the RTX 4080 while offering the same performance. I’m curious about what retailers will do with the existing RTX 4080 cards. Will they mark down the price, or will they jack up the prices of the RTX 4080 Super? I hope retailers will discount the old RTX 4080 instead.
GeForce RTX 4080 Super vs RTX 4080 Conclusion
TL;DR, the RTX 4080 Super has the same performance as the RTX 4080. Despite having slightly different specs, both cards perform effectively the same. There’s nothing “super” about the RTX 4080 Super’s performance. The real and only difference is the starting price.
Why did NVIDIA release a “new” GPU instead of discounting the older RTX 4080 to $999? Well, I think it’s marketing. Releasing a new product will create more buzz, reviewers will review the product, and it will get more attention.
There’s also a psychological aspect. Some people will prefer to have the newer product or better branding. Without knowing the performance difference between the RTX 4080 Super and RTX 4080, which will people likely pick? Most of the time, it would be the former.
Also, NVIDIA doesn’t usually discount their product. That’s AMD’s doing. NVIDIA would rather release something “new” and give it a new pricing. Well, except for the RTX 4070 Super and RTX 4070, since NVIDIA didn’t discontinue the RTX 4070. But between the two, the RTX 4070 Super is the better choice, if you’re interested to know.
At the end of the day, this is a boring release from NVIDIA. Among the RTX 40 Super GPUs, only the RTX 4070 Super was a bit exciting. The RTX 4070 Super offered a performance increase near the RTX 4070 Ti. And it also started at a lower price. Meanwhile, the RTX 4080 Super performs similarly to the GPU it replaces. But it comes with a lower starting price. Only the RTX 4070 Ti Super is the least exciting GPU since its performance is only around 5% to 7% faster, but the starting price is the same as the RTX 4070 Ti.