A two-generation gap, but how do they compare?
WePC is reader-supported. When you buy through links on our site, we may earn an affiliate commission. Prices subject to change. Learn more
Solid build quality and high finish
Excellent cooling solution to keep the temperature and performance up
Overclocked model gives you one of the top specs for the model
It is more expensive, pushing above the MSRP of the card
Great in-game performance
Very high build quality
Ray-tracing technology
Very efficient cooling
Few RTX-enabled games to enjoy
The third item in the Ada Lovelace roundup, the RTX 4070 Ti, brings the entry price down to an almost bearable $799. While DLSS 3 aims to improve the performance situation, raw performance is essentially on par with the Nvidia RTX 3090 but with half the memory and less power consumption.
At a lower price, it essentially ties the RTX 3090 and is faster than other products in the 30-series. This is profitable news. The unfortunate news is that this is Nvidia’s most costly xx70-class graphics card to date. Put another way, what we’re actually discussing feels more like an RTX 3080 12GB replacement.
| RTX 4070 Ti | RTX 2070 Super | |
|---|---|---|
| GPU | AD104-400 | TU104-410 |
| Shaders | 7,680 | 2,560 |
| RT Cores | 60 | 40 |
| TMUs | 240 | 160 |
| Memory | 12GB GDDR6X | 8GB GDDR6 |
| Base clock | 2,310 MHz | 1,605 MHz |
| Boost clock | 2,610 MHz | 1,770 MHz |
| Memory clock | 21 Gbps | 14 Gbps |
| Card bus | PCIe 4.0 x16 | PCIe 3.0 x16 |
| Memory bandwidth | 504.2 GB/s | 448.0 GB/s |
| TDP | 285 W | 215 W |
Technical comparison of the two graphics cards, with the Nvidia GeForce RTX 4070 Ti on the other side and the EVGA GeForce RTX 2070 SUPER on the other. The first one is for the RTX 2070 Super, which has 8GB of RAM and 2560 shading units with a base clock speed of 1610MHz. The RTX 4070 Ti is the second one. It has 12GB of memory, 7680 shading units, and a basic clock speed of 2310 MHz.
The only real question left is how performance scales down to the AD104 GPU with fewer shaders, less memory, less cache, a smaller memory interface, etc., since we’ve already covered Nvidia’s Ada Lovelace design. According to information that has leaked, the rebadged RTX 4080 12GB will be 20–25% slower than the RTX 4080, putting it in the RTX 3090 range.
The RTX 2080 Ti has around twice the performance of the RTX 2060, which was the slowest of the RTX 20-series GPUs. The RTX 3050 was three times as fast as the RTX 3090 Ti from the previous generation. Even if there is currently no RTX 4070 or RTX 4050, we may potentially see a performance increase of up to twofold from the larger, more expensive RTX 4090.
Great in-game performance
Very high build quality
Ray tracing technology
Very efficient cooling
Few RTX-enabled games to enjoy
A basic entry for the 4070 Ti that will cost close to MSRP
ZOTAC still has great quality builds even for a lower price
Will fall behind in perfromance without a factory overclock
We can compare the graphics card with their score on 3DMark. A straight comparison between the two will provide good information. And there we see the RTX 4070 Ti achieves a score of 22,652. Whilst the RTX 2070 Super only gets 10,119. This means the 4070 Ti is 124% better than the 2070 Super.
Now when it comes to the speeds there is also the frequency of the GPU that is a consideration. So the 2070 Super has a base clock of 1,605 MHz and a boost of 1,770 MHz. Whilst the 4070 Ti is capable of a base clock of 2,310 MHz and a boost clock of 2,610 MHz. That means the 4070 Ti has a 44% faster base clock and a 47% faster boost clock. So a significant improvement along with the architecture.
Trusted Source
WePC’s mission is to be the most trusted site in tech. Our editorial content is 100% independent and we put every product we review through a rigorous testing process before telling you exactly what we think. We won’t recommend anything we wouldn’t use ourselves. Read more
