AMD says their new Radeon RX 6800 XT graphics card competes with the Nvidia RTX 3080 for less money. I’m going to put this claim to the test by comparing both in 14 games at 4K, 1440p, and 1080p resolutions and content creator workloads to see what the differences are.
Let’s start with the spec differences, I think the main thing to note is that the 6800 XT has 16gb of memory, more than the 10 in the 3080, though the 3080’s memory is faster. The 6800 XT costs $50 less than the 3080 though, but it will also be interesting to see if they will have issues keeping it in stock like Nvidia has had.
Test PC Setup
This is the system that I’m using to test out both graphics cards, it’s got an overclocked i9-10900K at 5.2GHz with 32gb of DDR4-3200 CL14 memory in dual channel. This should keep things fair for each GPU, however Radeon 6000 series GPUs can get a boost with an AMD processor, but that’s a topic for another time. We’ll first dig into the gaming results, then check out cost per frame, power draw, and content creator workloads afterwards.
Microsoft Flight Simulator was tested in the Sydney landing challenge, and there’s a trend here that appeared throughout a lot of the games tested, where the 6800 XT was ahead at 1080p, but then the 3080 started pulling ahead as we move up to higher resolutions, despite the 6800 XT having a higher memory capacity. Red Dead Redemption 2 was tested using the games benchmark tool. This time the 3080 was just straight up ahead regardless of resolution, though not by large margins. The 6800 XT was 5% slower at 1080p, but then 12% slower at 4K, so again the 3080 is doing better at higher resolutions. Battlefield 5 was tested in campaign mode by running through the same mission on both graphics cards. Both were hitting the 200 FPS cap at 1080p so that’s not too useful. At 1440p the 6800 XT was reaching 14% higher average FPS when compared to the 3080, then at 4K both are doing basically the same, with a little more of an edge in 1% lows to the 6800 XT. Shadow of the Tomb Raider was tested with the games built in benchmark. This was another of the games where the 6800 XT was ahead at lower resolutions, but then at 4K the 3080 was 4% ahead, which in this case was just 3 FPS, so only small differences either way. Watch Dogs Legion was also tested with the games benchmark tool. The 6800 XT was 8% faster than the 3080 at 1080p, at 1440p they’re about the same, but then at 4K the 3080 was almost 10% faster in average FPS with a 7% higher 1% low. For Control I’ll start with ray tracing off results. The RTX 3080 was ahead at all resolutions in this one, and at 4K and 1440p resolutions the performance difference was the largest out of all titles that I’ve tested, but even so I still found 4K on the 6800 XT quite playable, and 100 FPS at 1440p is still a respectable result. I’ve tested ray tracing with the high profile enabled. No DLSS was used here, as that’s an Nvidia only feature, however that would also be an advantage to the Nvidia platform in the games that support it. Anyway the 3080 was much further ahead compared to most of the other games tested when looking at ray tracing, not too surprising I guess given it’s been made with the Nvidia implementation in mind as that’s what’s been available for two years now. Dirt 5 on the other hand was a massive win for the 6800 XT. Full disclosure, AMD provided a key for this game and I wasn’t sure if I’d add it in as I knew it was clearly going to favour them, but hey it is a new game people will play, and the ray tracing results are in contrast to what we just saw before in Control which I thought was worth discussing. As this title was developed for ray tracing on RDNA 2, we’re seeing pretty much the opposite of control with ray tracing enabled where Nvidia had an above average lead there, so basically it seems to depend on the specific game and how it’s optimized rather than just a blanket statement that one’s better than the other in ray tracing. I’ve tested Metro Exodus with the game's benchmark tool. This was another win for the 6800 XT at all resolutions, though the differences are much smaller this time, only one to two frames different in the averages so kind of margin of error stuff, though there was a bit more of a difference noted in 1% lows. The Witcher 3 had one of the smallest differences out of the games tested in average FPS, as you can see it’s quite close at 1080p and 1440p, though the difference at 4K was a little bigger, it’s less than 4 FPS. The 6800 XT was however notably better in 1% lows across all three resolutions. Speaking of small differences, death stranding was in a similar boat, at least in terms of average FPS. The 6800 XT was slightly ahead in the averages, however this time the 3080 had the advantage in 1% lows at 1440p and 1080p. Assassin’s Creed Valhalla was tested with the games benchmark tool. The 6800 XT had a much more obvious lead in this one, regardless of the resolution in use. It was 8% faster than the 3080 at 4K, but then at 1440p it’s now 21% faster, and then 29% faster at 1080p, the second largest difference out of the 14 games tested. Assassin’s Creed Odyssey was also tested with the built in benchmark, I tested this before Valhalla came out so figured I’d include it as I’ve got the data. The results here are apparently the opposite of many other games I’ve tested, where the 3080 was doing better at 1080p, but then at higher resolutions the 6800 XT was in front, with a 12% lead at 1440p and 7% at 4K. Call of Duty Modern Warfare was tested in campaign mode. Like a lot of other games, only minor differences here. Basically the same average FPS at 4K and 1440p, while the 6800 XT was almost 5% better at 1080p. Rainbow Six Siege was tested using the games benchmark tool with Vulkan. I feel like I’ve said this enough times already, but once more the 6800 XT was doing better in average frame rate at 1080p, but then the 3080 takes the lead at 1440p and 4K. The 3080 was just 2% faster at 1440p and 6.5% faster at 4K.
1080p Game Results
At 1080p on average over all 14 games tested the 6800 XT was 5% faster than the RTX 3080 in average FPS. Dirt 5 and Assassin’s Creed Valhalla are definitely boosting the overall average, but even if we remove these as outliers the 6800 XT was still slightly ahead on average.
1440p Game Results
Stepping up to 1440p and although the difference in Dirt 5 is still the biggest out of the games covered, the gap isn’t as large now when compared to some of the other titles. The 6800 XT is around 4.6% faster than the 3080 on average here.
4K Game Results
At 4K the 6800 XT was 1.4% slower on average than the 3080, and again without dirt 5 it was closer to being 3.4% slower, so a bit of a difference that one game can make, but at the end of the day there’s still not a huge difference between these graphics cards on average.
Which For Gaming?
On average though I don’t think the performance differences are enough to make a noticeable difference either way, as many of the games are performing quite closely together, though there are of course outliers. If you mainly play titles that do one way or the other, then it might make sense to pick one for say Dirt 5 or another for Control, but for the most part in the selection of games I’ve tested, it’s really difficult to go too wrong with either, as both offer great performance.
Cost Per Frame
So with performance being somewhat close together on average, the 6800 XT ends up offering better value as it’s $50 cheaper than the 3080, and this was the case at all three resolutions tested. It will be interesting to see if Nvidia does anything to counter this.
The power draw results were interesting. The system with the 3080 installed was only drawing a little more power in this test, 4% higher wattage in Control at 4K with RTX off, but if you recall the 3080 was also performing a fair bit better here with a 17% higher average frame rate in the same test, granted this game did see the best result on the 3080 out of the titles tested at 4K, it just happens to be the game I’ve been using to compare power draw with recently.
Performance Per Watt
In terms of performance per watt, when looking at the results from all 14 games, the 6800 XT is able to get more done with the same amount of power, as again on average the frame rates aren’t too different.
Content Creator Workloads
This might sound crazy, but there’s more than just gaming out there. I’ve also compared both in content creator workloads. I’ve tested DaVinci Resolve with the Puget Systems benchmark, and the 3080 was scoring around 27% higher than the 6800 XT, so a much bigger difference when compared to most of the games covered previously. I should note that the results on the 3080 were far more consistent, whereas I had to run the test 7 times on the 6800 XT to arrive at this average, as I was getting scores between 813 and 979. I’ve used Adobe Premiere to export one of my review videos at 4K, and the 3080 was faster in this one too, again the difference was more so when compared to most of the games. I’m not sure if both of these software suites need to be updated for the RX 6000 series though. Blender was tested with the Open Data benchmark. Interestingly the 3080 was only completing the classroom test a second faster than the 6800 XT, but in the shorter BMW test the 3080 was 46% faster, so quite a difference.
In general it would appear that the 3080 has the edge outside of gaming. So all things considered, the Radeon RX 6800 XT is looking like a compelling option. It’s coming in cheaper than the Nvidia RTX 3080, and in many of the games tested the performance differences just aren’t that big, though this does of course vary by game. Combined with the fact that the 6800 XT has more VRAM, I’m expecting it to last longer and so better at higher resolutions years from now, that will definitely be worth revisiting in future.
Driver Stability and Availability
The fact that AMD is even competing at RTX 3080 level now is pretty impressive, but it will be interesting to see how things like driver stability go over time, as that was a pretty big issue with Radeon graphics cards last generation. I did actually have some game crashes on the 6800 XT, both in red dead redemption 2 and metro exodus, but to be fair I had both of those same games crash on a recent Nvidia based laptop just last week.
I’m just a small sample size of one, we really need to wait and see what happens once people start getting these products into their hands and get them running in more and different hardware configurations. Likewise it’s also going to be interesting to see how much stock is available, if it’s hard to buy on launch for months like the RTX 3080 then that’s not going to be a great experience for customers.
In addition to the XT model, I’ll also be comparing the non XT version against other graphics cards.