NVIDIA knew from the RTX 4080's launch that pricing the 80-class card could mean another disaster. While it released the RTX 4080 Super later which one would expect to be much faster on paper, it came out to be a refreshed RTX 4080 with marginal improvements for $200 less.
Once again, NVIDIA is back with another 80-class card in the Blackwell lineup, i.e., the GeForce RTX 5080. The hopes for this card were high and even though we did see some gains over the RTX 4080 and 4080 Super, they weren't as impressive as we expected before the launch. Still, the RTX 5080 isn't just about gaming (as far as NVIDIA claims) and, as you would expect, NVIDIA wants you to leverage the RTX 5080's AI capabilities through its breakthrough Blackwell architecture.
If you have been wondering whether the GeForce RTX 5080 is a good upgrade over the RTX 4080/4080 Super (or over RTX 4070 cards), here you will find all your answers. From architecture to gaming benchmarks and technology advancements, here is everything you should know about what the RTX 5080 brings new to the table.
Latest Update: 19th October, 2025: Added benchmark charts comparing both GPUs head to head in multiple games at both 1440p and 4K resolution.
Quick Specs: RTX 5080 vs RTX 4080 Super
| Specs/GPU | RTX 5080 | RTX 4080 Super |
|---|---|---|
| Architecture | Blackwell | Ada Lovelace |
| GPU Die (Size) | GB203 (378 mm2) | AD103 (379 mm2) |
| Process Node | TSMC 5nm | TSMC 5nm |
| Cuda Cores | 10752 | 10240 |
| SMs | 84 | 80 |
| RT Cores | 84 | 80 |
| Tensor Cores | 336 | 320 |
| L2 Cache | 64 MB | 64 MB |
| FP32 | 56.28 TFLOPS | 52.22 TFLOPS |
| FP4 | 1801 TFLOPS | 836 TFLOPS |
| VRAM | 16 GB GDDR7 | 16 GB GDDR6X |
| Memory Bus | 256-bit | 256-bit |
| Memory Speed | 30 Gbps | 23 Gbps |
| Memory Bandwidth | 960 GB/s | 736.3 GB/s |
| Base/Boost Clock | 2295/2617 MHz | 2295/2550 MHz |
| PCI-E Interface | 5.0 | 4.0 |
| TDP | 360W | 320W |
| MSRP | $999 | $999 |
Architectural Differences
Transitioning from Ada Lovelace to Blackwell, the RTX 50 series brings several refinements over the predecessor. One of them is how the SMs (Streaming Multiprocessors) are structured. SMs are responsible for executing parallel tasks, and NVIDIA just combined both FP32 and INT32 cores to offer twice the throughput through INT32, allowing faster processing of mixed workloads.
On the Ada Lovelace, the SMs had a separate FP32 block of Shader Cores, responsible for executing general-purpose graphics and compute tasks, and another with a mixture of FP32 and INT32 cores. This small change is then followed by a Blackwell Shader Execution Reordering, which distributes the type of workload to either the Tensor Cores or to the Shader Cores (FP32/INT32).
Compared to the Ada cards, the Blackwell GPUs offer up to two times the SER, which is responsible for a noticeable increase in processing speed.
Apart from the core design changes, the RTX 5080 also brings the latest generation of RT and Tensor cores, including 4th Gen RT cores, which offer Triangle Cluster Intersection Engine, optimized for Mega Geometry. As per NVIDIA, the 4th Gen RT cores can reduce the memory footprint by 25% and also increase the Ray Triangle Intersection Rate by two times compared to the RTX 4080 Super, making the RTX 5080 (and RTX 50-series as a whole) more optimized for Ray Tracing tasks.
Plus, the 5th Gen Tensor Cores also help the RTX 5080 and other RTX 50 series GPUs to increase throughput by twice compared to the Ada cards. As far as the architectural changes are concerned, these are the major ones, but at the end of the day, all we care about is how much improvement these bring to the table in real-world applications.
RTX 5080 vs RTX 4080 Super: Changes in Specs
Unfortunately, the RTX 5080 isn't as impressive as its architectural changes when it comes to its specifications. If you have gone through the spec comparison table above, you would have already got the idea of where the RTX 5080 stands. Surely, architectural changes do enhance performance noticeably, but there are limits to them, which depend on the core specifications.
Almost every spec shows a slight increase over the RTX 4080 Super, but these are generally modest, staying under 10% most of the time. For instance, the Cuda Cores on the RTX 5080 increased by just 5%, and SM, RT Core, and Tensor Core counts didn't see much of an upgrade (except for the generation). However, a major difference can be seen in the FP4 performance, which has more than doubled compared to the RTX 4080 Super i.e., 1801 TFLOPS vs 836 TFLOPS.
FP4 is specialized in accelerating AI computations as it can reduce the data size and computational load. This is one of the reasons why NVIDIA markets its RTX 50 series GPUs for not just gaming but for executing AI operations as well. As the RTX 5080 brings the FP4 format natively, it makes it hugely advantageous over its predecessor.
But what about gamers? It is still obvious that NVIDIA didn't go all in to offer anything significant for gaming. The RTX 5080 still retains the same VRAM capacity as the RTX 4080 Super, as well as the same memory bus. Except for the GDDR7 memory, there is nothing else that is helping the RTX 5080 gain a big advantage in VRAM configuration. 16 GB is surely enough for most titles, but at $1000, I don't think this is something that we can appreciate.
Upscaling Technology
Since we have NVIDIA, AMD, and Intel bringing enhanced versions of their proprietary upscaling technologies, there is no reason why it should be ignored in comparison. The RTX 5080 Super supports the latest DLSS 4 (Deep Learning Super Sampling), and while the RTX 4080 Super also supports DLSS 4, there are a few differences.
With the DLSS 4 launch, NVIDIA also introduced the Multi-Frame Generation technique. This is different from the classic Frame Generation and is only supported on the RTX 50 series currently. Instead of introducing a single frame between two frames, the MFG adds three fake frames for more smoothness. On paper, MFG can increase the FPS by up to 2X easily compared to Frame Generation, and it might come to the RTX 40 series as well.
Except for the MFG, both the RTX 40 series and RTX 50 series got the latest Transformer model for Super Resolution and Ray Reconstruction. Therefore, there is not much difference here between the two and the RTX 5080 can only have the advantage of MFG at the moment.
Synthetic And Gaming Performance
We have already published a couple of RTX 5080 reviews. These include the synthetic as well as gaming benchmarks to show you how fast the new 80-class card is against the RTX 4080 Super. Here is a quick summary of what you should expect with the RTX 5080.
In Synthetic tests such as FireStrike, Time Spy, Port Royal, and others, the RTX 5080 performs just fine in comparison to the RTX 4080 Super. However, in some cases, the performance difference isn't satisfactory. Overall, you are going to see 8-15% uplifts over the RTX 4080 Super. Comparing this to what the RTX 4080/Super brought over the RTX 3080, the RTX 5080 does disappoint a bit, particularly because it is still far behind the RTX 4090.
GeForce RTX 5080 vs RTX 4080 Super (FPS) @1440p Max Settings
The same is repeated in gaming benchmarks, which is why the RTX 5080 faces criticism. Unlike how NVIDIA portrays the RTX 5080 as being nearly twice as powerful as the RTX 4080, the reality is that it can't achieve such performance without MFG. In our testing, we found that the RTX 5080 is just 10-15% faster than the RTX 4080 and 4080 Super with and without Ray Tracing. However, there are some titles where we can see some good gap, particularly in the Resident Evil Village title where the RTX 5080 outperforms the RTX 4080S by around 26% at 1440p and by nearly 30% at 4K.
Now, even though it looks decent considering that there hasn't been a price increase over the last gen, this is perhaps one of the lowest generational uplifts we have ever seen on an 80-class card. Usually, 80-class cards would outperform the 90-class GPUs from the previous generation, but it's not even close this time.
GeForce RTX 5080 vs RTX 4080 Super (FPS) @4K Max Settings
Power Consumption
We are seeing a noticeably higher TDP rating for the GeForce RTX 5080 compared to the RTX 4080, but the default TDP doesn't tell us the whole story. For instance, even though the RTX 5080 is rated at 360W, it doesn't mean that it will consume 360W all the time. The total power consumption depends on the load and, as per our testing, we found that the RTX 5080 is slightly more power-hungry than the RTX 4080.
In gaming loads, the RTX 5080 can consume around 300W of power, while the RTX 4080 Super is also close, but expect 10-15W lower. AIB editions usually increase the power consumption a bit, but it's not much unless you manually overclock the GPU. In synthetic benchmarks, assuming the GPUs are maxed out, the RTX 5080 can consume up to 328W while the RTX 4080 Super maxes out at 315W.
Once again, not much difference. Considering we have a 10-15% performance difference between the two GPUs, the RTX 5080 is a bit better in terms of performance per watt than the RTX 4080 Super, which is appreciable. But we are more impressed by its thermals, and we have seen that AIB editions of RTX 5080 do even better.
Price and Value Comparison
Here is the ultimate deciding factor that reveals who's the boss. Taking everything above into consideration, the RTX 5080 comes out to be better than the RTX 4080 Super. This is true even when we consider its price. Both the cards were launched at $999 and the RTX 5080 offers better value, whether we talk about synthetic performance, gaming tests, thermals, and upscaling technology. Keep in mind that there are RTX 5080 editions that are still selling for over $1200, which negates any benefits it brings over the RTX 5080 Super if directly compare the performance per watt.
The Verdict
The GeForce RTX 5080 does outperform the RTX 4080 Super in most areas, but there are two key points to keep in mind: Number One, is that the RTX 5080 is a successor to the RTX 4080 Super, and we have to consider it a major factor when comparing the two. Number Two is how much value for money you are really getting when upgrading from the RTX 4080 Super to the 5080.
As far as the latter is concerned, no sane person should ever think about swapping an RTX 4080 Super with an RTX 5080. This is simply not logical or beneficial. Now, considering the first point, we can see that the RTX 5080 does offer somewhat better value than the RTX 4080 for the same price. However, the generational uplifts are disappointing, but at the same time, if you are upgrading from a significantly slower GPU, let's say an RTX 4060 or an RTX 3070, there is no reason why anyone would prefer the RTX 4080 Super over the RTX 5080, considering the latter is already discontinued.
To conclude, the NVIDIA GeForce RTX 5080 is great if you are building a brand new high-end gaming PC from scratch or upgrading from a budget or mid-range GPU. On the flip side, it's discouraged to upgrade from something like an RTX 4080/Super or RX 7900 XTX as you would be burning your money for small gains.
Follow Wccftech on Google to get more of our news coverage in your feeds.
