The RTX 3080 was once one of the most powerful cards you could buy – a must-have for all PC gamers looking to sport the best hardware. But now, with two variants, a 10GB and the later released 12GB version, the 3080 is very much middle of the road. So how does it compare with one of Nvidia’s newer cards, RTX 4070 Super? Can Nvidia’s once third-generation halo card keep up with today’s upper middle-weight contender in the 4070 Super?
In this feature we’re comparing both the new mid-range offering with the previous-generation high-end champion to see which one deserves your time and money. What’s more, we’re also bringing you our dedicated RTX 4070 Super review that goes into even more depth.
RTX 4070 Super vs RTX 3080 – specs
The differences between the two graphics cards are apparent throughout their specifications. The RTX 4070 Super comes with the recent Ada Lovelace architecture, while the RTX 3080 is built on the reliable Ampere architecture. The processor size is also a notable difference, with the RTX 4070 Super sporting a smaller 5 nm size compared to the RTX 3080’s 8 nm. Both cards utilize GDDR6X memory, but the RTX 4070 Super edges ahead with 12 GB of VRAM compared to the RTX 3080’s 10 GB. However, the RTX 3080 strikes back with a higher memory bandwidth of 760.3 GB/s, leaving the RTX 4070 Super trailing at 504.2 GB/s.
| Specifications | RTX 4070 Super | RTX 3080 |
|---|---|---|
| Architecture | Ada Lovelace | Ampere |
| Processor Size | 5 nm | 8 nm |
| Transistors | 35,800 million | 28,300 million |
| Base Clock | 1980 MHz | 1440 MHz |
| Boost Clock | 2475 MHz | 1710 MHz |
| Memory Clock | 1313 MHz | 1188 MHz |
| Memory Type | GDDR6X | GDDR6X |
| Memory Size | 12 GB | 10 GB |
| Bandwidth | 504.2 GB/s | 760.3 GB/s |
| TDP | 220 W | 320 W |
| Shading Units | 7,168 | 8,704 |
| L1 Cache | 128 KB (per SM) | 128 KB (per SM) |
| L2 Cache | 48 MB (shared) | 5 MB |
| Price | $599 | $699 |
When it comes to clock speeds, the RTX 4070 Super takes the lead with a base clock of 1980 MHz and a boost clock of 2475 MHz; surpassing the RTX 3080’s 1440 MHz base and 1710 MHz boost clocks. In terms of transistors, the RTX 4070 Super boasts 35,800 million, outshining the RTX 3080’s 28,300 million. In the power consumption department, the RTX 4070 Super demonstrates efficiency with a TDP of 220 W, while the RTX 3080 draws more power with a TDP of 320 W. Cache sizes, including L1 and L2, exhibit subtle differences but contribute to the overall performance variation between the two cards. So what does this mean in the real world? Utimately, the 4070 Super is moe power efficient, so will cost you less per year. It’s also more powerful and with its superior VRAM gives you more headroom to run games at higher settings and more demanding resolutions. Let’s take a closer look at the performance.
RTX 4070 Super vs RTX 3080 – performance
The RTX 4070 Super, with its superior clock speeds and transistor count, showcases impressive capabilities. The 7,168 shading units contribute to enhanced rendering and graphics processing. However, the RTX 3080, with its higher memory bandwidth, excels in scenarios that demand rapid data access. RTX 4070 Super significantly boosts gaming and creative applications, providing smoother gameplay experiences and faster rendering times. The RTX 3080, on the other hand, remains a formidable competitor, particularly in memory-intensive tasks. A YouTube video from the Mark PC channel runs different games using the two GPUs to see the difference. The result shows that both GPUs have minimal performance difference, with RTX 4070 Super having the upper hand, especially in high–quality graphics. This will likely be due to both the increased VRAM and the improved memory clock when comparing the Super with the 3080.
Out with the old, in with the new
The 4070 Super is a fantastic card, so don’t let it ‘mid range’ tag affect you. The difference between the RTX 3080 is marginal, and for a slightly higher cost you’re getting a card with Ada architecture and a decent increase in VRAM. If you do decide to get a 3080, opt for a 12GB version – which at least gives you more headrooom, although we advise sticking with the 40 Series.
Of course, as is the case with Nvidia GeForce RTX cards, both cards support ray tracing and DLSS. This means graphically intensive games such as Cyberpunk 2077, Alan Wake 2 and Hogwarts Legacy will run pretty well on both cards. It’s worth noting that the 10GB 3080 will almost certainly need some of the settings turned down if you’re looking for plus 60FPS in more demanding games.
RTX 4070 Super vs RTX 3080 – price and value
As price often plays a crucial role in consumer decisions, the RTX 4070 Super currently retails for around $599. The RTX 3080, now a couple of years old, can be picked up for less than this, with some models on eBay closer to the $400 mark. This is a far cry from the initial launch price, which topped more than $1000. In our opinion, if evaluating on price alone, we would always opt to pay a little more for the newer hardware. Not only are you getting the latest Ada GPU architecture making this a worthwhile upgrade, but you’re also future-proofing yourself thanks to more visual memory. The choice between the two boils down to individual preferences and budget constraints. Considering the price difference, the RTX 4070 Super delivers excellent value for gamers and content creators seeking high-quality performance without breaking the bank.
Final verdict
Each graphics card has its strengths and weaknesses in the RTX 4070 Super vs RTX 3080 showdown. The RTX 4070 Super impresses with its advanced architecture, competitive specs, and attractive pricing. The RTX 3080 was once a formidable option and the MSRP matched that, but in the advent of Nvidia’s Super cards, the 3080 quickly starts to not make sense for those looking for a serious upgrade in 2024. If you’re looking for other graphics cards, see the best of what Team Red and Team Green have to offer in our best graphics cards buying guide.
