ASUS GeForce RTX 3060 TI ROG Strix Gaming OC
Pros
Excellent aesthetic design
Great cooling solution
High boost clock speed
Cons
On the higher end of the price spectrum
Nvidia GTX 1080 Ti vs RTX 3060 Ti
Can the Pascal giant hold up against the Ampere lower end?
Updated: Nov 22, 2022 3:05 pm
Share this article...
WePC is reader-supported. When you buy through links on our site, we may earn an affiliate commission. Prices subject to change. Learn more
Table of Contents
hide
With the 30 series complete there are still comparisons you may be wondering about. So this time we look at how well the old holds up against the new with the 1080 Ti vs 3060 Ti. A two-generational difference there are a lot of differences to expect, especially with the change from GTX to RTX and seeing if the top end of Pascal holds up to the lower end Ampere.
Of course, there are some big changes between these generations so we expect a few key improvements and potential performance enhancements. The 1080 Ti came out in March 2017, whilst the 3060 Ti was over three years later. Releasing in the December of 2020, slightly later than the initial 30 series release.
ASUS ROG STRIX GTX 1080 Ti
Pros
Strong factory overclock, but not the best
Best cooling performance of the bunch offers more OC headroom, plus…
Additional features like RGB lighting for extra visual flair
Cons
It’s also huge, but surprisingly not the biggest
It’s fairly expensive
Architecture
Obviously one of the core upgrades will be the architecture of the two graphics cards. It is the basis of how they work and create the graphics you see. One key feature difference is the introduction of RT cores.
Looking into it, we see the GTX 1080 Ti feature a GP102-350 GPU variant. Based on the Pascal architecture it is created with TSMCs 16nm process. It is 471mm² in size and made with 11.8 billion transistors. Whilst the 3060 Ti is the Ampere GA104-200 GPU die. Made with Samsungs 8nm process it creates a 392mm² die with 17.4 billion transistors.
Specifications
| GTX 1080 Ti | RTX 3060 Ti | |
|---|---|---|
| GPU | GP102-350 | GA104-200 |
| CUDA cores | 3,584 | 4,864 |
| RT cores | 0 | 38 |
| TMUs | 224 | 152 |
| Memory | 11GB GDDR5X | 8GB GDDR6 |
| Base clock | 1,481MHz | 1,410MHz |
| Boost clock | 1,582MHz | 1,665MHz |
| Memory clock | 11 Gbps | 14 Gbps |
| Card bus | PCIe 3.0 x16 | PCIe 4.0 x16 |
| Memory bandwidth | 484.4 GB/s | 448 GB/s |
| TDP/TBP | 250W | 200W |
There is also the rest of the specifications of the graphics card to consider. Although the architecture may vary there are still some other parts that are more standard. So we do take a look at what they may offer between them.
GPU die cores
Die cores are what make up the GPU chip itself. Each is specialized in what they work on. Working together to create the final graphics scenes to display. In particular, we have the shaders that create shadows and shade in the scene. Along with RT cores calculating the real-time ray tracing to add to the graphics and TMUs creating the 3D models from bitmaps.
Between the two cards, we see some differences. In terms of CUDA cores, the 1080 Ti has 3,584 compared to the 3060 Tis 4,864. Along with likely optimizations down the road. In TMUs there is a decrease, down from 224 to 152. From a possible optimization or moving the work away from them or it just has less. Whilst in RT cores the Pascal card has none. The option only became available with RTX and the Ampere has 38. So to do any ray tracing you need that option.
Video memory
VRAM is the storage for your display’s image data once it has been processed. So used as a buffer, it enables the smooth experience of your graphics with framerates and latency. It’s especially important for demanding 3D graphics and higher resolutions that have more pixels to render.
There are some differences a few years apart. The 10 series card has 11 GB of GDDR5X, whilst the 30 series card has 8GB of GDDR6. Losing the capacity, but with an improved version. Which in fact is faster with 14 Gbps compared to 11. But since the 3060 Ti uses a smaller memory bus and so has a slightly lower bandwidth of 448 GB/s instead of 484.4 GB/s.
Frequency
The frequency or clock rates are how quickly the cores process instructions and create their graphics. It is a measure of how fast the silicon crystals flip in the chip which processes the graphics. Although not the only factor of performance it is subject to overclocking on the consumer end to get the most out of them.
In terms of 1080 Ti vs 3060 Ti, there is a mix of leads between them. The 1080 Ti has a faster base clock of 1,481 MHz compared to 1,410 MHz. Whilst the 3060 Ti has a faster boost clock reaching 1,665 MHz instead of 1,582 MHz.
TDP
TDP is the thermal design power of a graphics card. It is how much energy the card will consume at max utilization. Meaning how much you’ll be paying to run it and how much heat it will give off into the system. Between the two of the choices, it is in fact the older card with a higher rate. With a 50W higher count at 250W instead of the more efficient newer card with 200W.
PCIe bus
The card bus is the interface between the graphics card and the motherboard. It is how the parts communicate and work together. PCIe standard is set by SIGGRAPH and the generations have different speeds and capabilities.
So between generations, there has been a change. The 10 series features PCIe 3.0 x16 whilst the 30 series has PCIe 4.0 x16. This doubles the available bandwidth to the card from 15.754 GB/s to 31.508 GB/s. So it does allow for a greater performance available for the cards.
1080 Ti vs 3060 Ti performance
For looking at the performance we go to 3DMark, we get a comparative look at the graphical performance of the two. There we see the 3060 Ti achieve a score of 11,700 with a price-to-performance ratio of 29. Whilst the 1080 Ti only gets 9,899 with a ratio of 14. Showing a much better improvement over the old.
In terms of gaming performance, we look to reviews. In particular, we see Gamers Nexus review with gaming across the board. There at 4k in SOTTR the 3060 Ti leads by 18%, whilst at 1440p and 1080p that lead is only 15%. Although the same lead may not translate across all applications the 3060 Ti does lead most of the way. But at 4k the 1080 Ti with its bigger VRAM can lead in some games.
1080 Ti vs 3060 Ti price
Lastly, we look at the MSRP and asking prices of the card. As the 1080 Ti is EOL it’s unlikely you can find it for sale at the moment new. So we look at what their asking prices were to see their worth.
So when it first came out the GTX 1080 Ti had an MSRP of $699USD/£699. Whilst the RTX 3060 Ti is rather cheap in comparison. It releasing with a price tag of $399USD/£369. So you get a lot better price and performance with that option.
MSI GEFORCE RTX 3060 TI VENTUS 3X OC
Pros
Stylish, gaming aesthetics
Strong materials and build design
High-end cooling solution
Cons
On the expensive side
EVGA GTX 1080 Ti SC Black Edition
Pros
Best out-of-box performance thanks to a stellar factory overclock
Strong cooling performance provides even more OC headroom
Cons
It’s huge, even larger than the STRIX somehow
It’s pretty expensive
Is a 3060 Ti better than a 1080 Ti?
Overall, with a 1080 Ti vs 3060 Ti, the RTX card is a much better choice. With outright better performance in most situations and for a lower price it is an easy pick. Especially alongside improvements in architecture and features allowing it to utilize ray tracing and DLSS. Whilst on top is much more easily available.
Although if you’re considering upgrading at the moment it might be worth holding out. There is an expectation of the RTX 4000 series and RDNA 3 launching in 2022. Meaning even better performance and choices or falling prices to benefit from.
