The RTX 2070 launch date is set for October 17th and the retail price for the base model is $499 which is a $120 premium over the GTX1070 but the Founders Edition will cost you $ 599. Naturally, the AIB units will cost more. October is also when Microsoft will release the Windows 10 update which will include ray-tracing support.
The RTX 2070 is based on the TU106 silicon. Nvidia calls this as ‘The cheapest Turing card with ray tracing support’. Of course, there will be a debate about the RTX 2070 and its performance not to mention the lack of NV link.
Well, Nvidia does create trust issues with its customers but this is like any other company from time to time! Some issues are understandable once it’s explained properly with no marketing lingo which rarely happens and reviewers decrypt and scrutinize details or the lack of it!
Speaking of which this is related to those who have pre-ordered RTX 2080TI Founders Edition, the graphics card was supposed to be shipped on the launch day. But the delivery was delayed to September 27th and guess what? Now it’s delayed to be shipped between October 5th and 9th!
[amazon_link asins=’B07GHVWMBS,B07GHXQTGZ,B0762Q49NV’ template=’ProductCarousel’ store=’lmtorres-20′ marketplace=’US’ link_id=’94fc9e25-c4b8-11e8-84bc-11f3dd4f4b84′]
Understandably people who took Nvidia for their word without knowing its performance capability until the embargo was lifted people are pissed off! People who pre-ordered the Founders Edition of the RTX 2080Ti are pissed off! There’s no word why this is happening nor there’s any word about the availability of RTX 2080TI from its respective AIB partners.
The Nvidia RTX 2080 and GTX 1080 are both powerful gaming graphics cards, but which should you pick, and is it worth upgrading? In this article, we’ll check out a heap of game benchmarks at 4K, 1440p, and 1080p at all setting levels to help you see the performance difference between them. For the testing I’m using the MSI 2080 Duke, which actually has a little overclock out of the box, I’ve got a full review on that if you’re after more information.
The GTX G1080, on the other hand, is my EVGA FTW2, which also has a slight overclock over the reference specs. Let’s just quickly take a look at how the 2080 and 1080 differ in terms of specs, note that things like clock speed and power will vary between specific cards, these are just the reference specs for each model, with the speeds of my specific cards in brackets. Important differences to take note of are that the RTX 2080 has more Cuda cores but a lower base and boost clock speeds out of the box.
They both have the same sized 8GB memory, although the 2080 has the newer GDDR6 memory which is faster. Just before we get into the gaming benchmarks let’s check out the system that was used for testing. I’ve got an Intel i7-8700K CPU overclocked to 5.1GHz on all cores in order to best attempt to remove CPU bottlenecks, as well as 16GB of memory running at DDR4-2666 in dual channel.
All tests were run with the same Windows and game updates applied, however different Nvidia drivers were used, as the RTX cards launched with a newer driver, and I retested my 1080 card a few days before the release of the drivers. Given I’ve already compared the 2080 against the 1080 Ti and found that both were fairly similar, I expect the 2080 to come out a fair bit ahead.
With all of this in mind let’s check out the gaming results! Fortnite was tested with the same replay, at 1080p we’re getting excellent frame rates with either card, although the 2080 was performing 25% better at epic settings, a pretty big difference.
Moving up to 1440p the frame rates drop down a bit, but still a pretty good performance for the most part. The 2080 is still performing 24% better at epic settings in this test. At 4K the frame rates at higher settings drop down quite a bit, although with the 2080 we’re still able to average 60 FPS at epic settings, and the 2080 was around 23% better than the 1080 now.
Shadow of the Tomb Raider was tested with the built-in benchmark, at 1080p the 2080 moves further ahead of the 1080 as we step up in settings, at the highest settings the 2080 is 25% ahead of the 1080.
At 1440p the frame rates are still pretty decent with both cards, but there’s a larger 30% difference between the two now, although the 1080 was still able to average over 60 FPS here. At 4K you’ll probably need to use lower settings for a good experience.
At the highest settings the 2080 is now 35% ahead of the 1080, and a much higher 46% improvement at the lowest settings, however, keep in mind that there will be newer drivers for this game as it’s new.
Assassin’s Creed Origins was also tested with the built-in benchmark, the results were pretty good for this intensive test, and the 2080 was 13% ahead of the 1080 at max settings. Stepping up to 1440p the results are still pretty good for this test, and the 2080 is now performing 25% better than the 1080 in terms of average frame rates at max settings, with the 1% low of the 2080 almost ahead of the 1080’s average. At 4K the lower settings are performing quite a bit better than the higher ones, and at max settings, the 2080 was now scoring 32% ahead of the 1080. Far Cry 5 was again tested using the built-in benchmark, and at 1080p we’re seeing the smallest differences between the two cards out of the games tested so far, just a 9% boost with the 2080 at ultra settings.
At 1440p this rises quite a lot, with the 2080 now 23% ahead of the 1080 with ultra settings, the same 23% increase was also seen with low settings. At 4K the difference between the two cards rises further still, with the 2080 now performing 28% better than the 1080 at ultra settings, and the 1% lows of the 2080 are ahead of the 1080’s averages at all setting levels.
Ghost Recon was once again tested with the built-in benchmark, and as a more intensive game, the frame rates at ultra settings aren’t that high, even at 1080p, however, the 2080 was still 22% ahead of the 1080. At 1440p the 2080 is still able to average 60 FPS at ultra settings in this test and was now performing 23% better than the 1080, again with 1% lows above what the 1080 could average. At 4K the difference between the two rises further, with the 2080 now 27% ahead of the 1080 at ultra settings, although the frame rates are fairly low now and lower settings would definitely be needed to play well. Overwatch was tested in the training area while running through the same route, at 1080p the results are very nice.
As the game caps out at 300 FPS we’re able to hit this at most setting levels, although at max settings the 2080 was 17% ahead of the 1080. Going up to 1440p we saw the same 17% increase with the 2080 over the 1080 at epic settings, still extremely high results and running with no problems. At 4K the difference between both cards at max settings was again 17%, same as the other settings tested, and the frame rates with both cards were still decent, definitely playable at 4K, it felt smooth as even the 1080 had 1% lows above 60 FPS. CS:GO was tested with the Ulletical benchmark, and at 1080p the frame rates regardless of the setting level were very high, with the 2080 10% ahead of the 1080 with all settings maxed out.
At 1440p this increases to a 15% improvement with the 2080 at max settings, but again the results from both cards are again fairly high in this test. At 4K the frame rates are still quite high, and at max settings, the 2080 pushes further ahead of the 1080 at 17.5% now.
Rainbow Six Siege was tested with the built-in benchmark, and we’re seeing very high frame rates here, especially with the 2080, which was a large 32% improvement over the 1080 at ultra settings at 1080p. At 1440p the frame rates are still very high, even at max settings, but the 2080 is now 41% ahead of the 1080 at ultra settings. At 4K the 1% lows with the 2080 are ahead of all the 1080’s results, with the 2080 now a high 47% better in terms of average frame rates than the 1080 at ultra settings.
[amazon_link asins=’B07GG9L5X1,B07GJ3Z8W2,B07GPW17YM’ template=’ProductCarousel’ store=’lmtorres-20′ marketplace=’US’ link_id=’c4c63d63-c4b9-11e8-bb3c-67bf9f0d9abb’]
PUBG was tested using the replay feature, at 1080p the frame rates are quite high, considering this is a pretty unoptimized game, and there was a decent 29% improvement with the 2080 at ultra settings. At 1440p this rises a fair bit, with the 2080 now averaging 40% higher frame rates than the 1080 at ultra settings, and the frame rates at all settings with either card are decent for this game. At 4K the 2080 is still quite ahead of the 1080 with the 1% lows of the 2080 above the averages of the 1080, with the 2080 being 36% higher in average frame rates at ultra settings.
Shadow of war was tested using the built-in benchmark, and at 1080p there was a 28% improvement when maxed out using the 2080.
At 1440p the 2080 is again ahead in all tests, rising up to 36% higher than the 1080 at ultra settings at this resolution. At 4K the performance of the 2080 rises further ahead, up to 40% better than the 1080 now, and still fairly high frame rates at this high resolution at the lower setting levels. Watchdogs 2 is quite a resource intensive game, and there were very minimal differences at 1080p as we appear to be CPU bottlenecked for the most part, with just a small 2% improvement with the 2080 at ultra settings.
At 1440p the differences between the two rise quite a lot at the higher setting levels, with the 2080 performing 35% better than the 1080 at ultra, although minimal differences at the lower settings. At 4K the 2080 is still performing 30% better over the 1080 at ultra settings, and as we saw in other games even the 1% lows with the 2080 were ahead of the 1080’s averages. I’ve also tested some benchmark tools, in Unigine’s Superposition benchmark the 2080 was ahead in all tests, with the difference lowering at the higher levels. In the various 3DMark benchmarks the 2080 was again always ahead, although some test were closer together than others.
In terms of improvement, overall the games tested at max settings with a 1080p resolution on average the 2080 was performing 19.6% better than the 1080 between the games tested. At 1440p on average over the same games the 2080 was now performing 28.4% better than the 1080 in these titles, but again it really depends on the specific game. At 4K the 2080 was on average 30.2% better than the 1080, a similar improvement that I saw when comparing the 1080 against the 1080 Ti previously, which makes sense as the 2080 performs very closely to the 1080 Ti.
The performance difference between the two cards rises on average at the higher setting levels as there’s less chance of a CPU bottleneck and the differences between the GPUs starts to shine. It’s also worth noting the differences in my actual cards that I’m testing with, both my 2080 and 1080 are overclocked out of the box, as previously mentioned, so this will slightly affect the results, although it would be quite negligible, either way, the 2080 is way ahead of the 1080 and that fact won’t change regardless of the specific card in use.
If you want to see overclocking results check out the full review of the MSI 2080 Duke, link in the description. I also haven’t actually been able to test any RTX specific titles or settings, as there aren’t any at the time of recording, so once RTX is actually utilized more in the future, the 20 series may become more useful, but I suppose the testing here makes the comparison more apples to apples, as we’re testing with the same game settings. Now for the final difference, the price. I suggest checking updated prices using the links in the description, as they will change over time. My 2080 Duke from MSI that I’m testing with is around $830 USD at the time of recording, but the 2080 prices seem to range from $750 to $900.
The 1080, on the other hand, can be picked up for around $450 USD to around six or seven hundred for a higher end model, and that’s forgetting the second-hand market which has a lot of deals at the moment due to ex-miners selling off cards. So which graphics card would you pick? The older but still capable GTX 1080, or newer RTX 2080?
Although the 2080 was a fair bit ahead of the 1080 in these games, the 1080 Ti still performs closely to the 2080, as I found out in my 2080 vs 1080 Ti video, link in the description. In that video, I noted that the 1080 Ti was cheaper than the 2080 meaning the 1080 Ti was offering better value, though this could of course change in the future. It pretty much seems like in the 20 series the tiers have just been shuffled over one spot, as the 1080 Ti is quite similar to the 2080, I suspect the 1080 will perform close to the soon to be released 2070, but we’ll have to test that in a future video once it’s released.
In any case I think it’s more likely someone would be looking at the performance differences between the 1080 and 2080 from the perspective of already upgrading from their 1080, in which case I’d suggest looking at the 1080 Ti instead as the 2080 is just so expensive at the moment, unless you really want to take the gamble today on ray tracing and DLSS technologies being the future.