cancel
Showing results for 
Search instead for 
Did you mean: 

RX 5700 (XT) vs. RTX 2070 & 2060 (Super) vs. GTX 1080 (Ti) vs. RTX 2080 (Ti)

RuneSR2
Grand Champion
In 1920x1080 we got the following average scores based on more than 20 games:



Or we could cut the cake like this - keeping everything relative to the 2080 Ti:

RTX 2080 Ti = 100 %
RTX 2080 =  86 %
RTX 2070 Super = 81 %
GTX 1080 Ti = 79 %
RX 5700 XT = 74 %
RTX 2070 = 73 %
RTX 2060 Super = 70 %
RX 5700 = 65 %
GTX 1080 = 63 %
RTX 2060 = 63 %

Or RTX 2080 Ti is 60 % faster than both GTX 1080 and RXT 2060...

In 4K we've got:



Or keeping everything relative to the 2080 Ti:

RTX 2080 Ti = 100 %
RTX 2080 =  79 %
RTX 2070 Super = 73 %
GTX 1080 Ti = 72 %
RX 5700 XT = 64 %
RTX 2070 = 64 %
RTX 2060 Super = 62 %
RX 5700 = 56 %
GTX 1080 = 54 %
RTX 2060 = 53 %

Or RTX 2080 Ti is 86 % faster than GTX 1080 and 88 % faster than RXT 2060, but still not fast enough to ensure that 45 fps are turned into 90 fps if you've got a GTX 1080 or RTX 2060. 

Source: https://www.techpowerup.com/review/amd-radeon-rx-5700-xt/

Oculus Rift CV1, Valve Index & PSVR2, Asus Strix OC RTX™ 3090, i9-10900K (5.3Ghz), 32GB 3200MHz, 16TB SSD
"Ask not what VR can do for you, but what you can do for VR"

16 REPLIES 16

Anonymous
Not applicable
I'd love to own an AMD GPU again!

RuneSR2
Grand Champion
Had the Hercules 3D Prophet Radeon 9700 Pro (fastest card for Doom3 back then) and HIS X800XT IceQ2 (fastest and most silent card for HL2 back then), but always preferred Nvidia drivers. But yeah, if the performance, noise levels and power consumption are right...

Man, is that really 15 years ago, sigh... And we're still waiting for HL3... Still love that box, here from 9700 non-pro, but the pro box was similar 🙂
plxneq55djqh.jpgdrwf34nvcvfm.jpg

For those who have long forgotten the old Doom3 king:


- and the king of Half-Life 2 at launch:
Image result for x800xt ice q 2

BTW, this is an image showing how much ATi dominated with the Radeon X800XT PE, when Half-Life 2 arrived:



The scores are average scores (Radeon X800XT set to 100%) based on Half-Life 2 benchmarks using 1600x1200 + 4x antialiasing + 8x anisotropic filtering (results made and published by Firingsquad, XBitLabs, DriverHeaven and more). 
I think this was the last time Radeon really dominated, and it may take a tremendous amount of effort for AMD to get Radeon back on the throne. These results are from 2004. 

Oculus Rift CV1, Valve Index & PSVR2, Asus Strix OC RTX™ 3090, i9-10900K (5.3Ghz), 32GB 3200MHz, 16TB SSD
"Ask not what VR can do for you, but what you can do for VR"

Anonymous
Not applicable
I had a Powercolor HD4870 2GB for YEARS. Really did punch above its weight due to having twice as much VRAM compared to most other cards at the time.

Techy111
MVP
MVP
HD7970 eased me into VR with the DK2 until I got a 980 😉
A PC with lots of gadgets inside and a thing to see in 3D that you put on your head.

hoppingbunny123
Rising Star
you want to compare pixel fill rate, see video details for more details;


GTX 1060 6GB = Memory 6144 MB, GDDR5, Bandwidth 196608 MB/sec, Power (Max TDP) 120 watts

RX 5700 (XT) = Memory 8096 MB, GDDR6, Bandwidth 458752 MB/sec, Power (Max TDP) 225 watts

RTX 2060 super = Memory 8192 MB, GDDR6, Bandwidth 458752 MB/sec, Power (Max TDP) 175 watts

RTX 2070 super = Memory 8192 MB,  GDDR6, Bandwidth 458752 MB/sec, Power (Max TDP) 215 watts

GTX 1080 (Ti) = Memory 11264 MB, GDDR5X, Bandwidth 495616 MB/sec, Power (Max TDP) 250 watts

RTX 2080 (Ti) = Memory 11264 MB, GDDR6, Bandwidth 630784 MB/sec, Power (Max TDP) 250 watts


you can see here the 2080ti has more ram, more bandwidth, and more power.

i find after watching msi afterburner the gpu used is equivalent to the power used, not exactly but theres a close equivalence, so the power draw on the videocard relates to the power the gpu can work at.

then note the low power of the 2060 super, but because it has 100MB more memory than the 5700xt it performs better than it would if it had the same amount of memory. meaning the only real difference between the 2070 super and 2060 super is the amount of power used which relates to this article;

note the temperature difference between the 2060 super and 2070;


2060 super = 72C, 2070 = 
79C

they show the 2070 is being converted to the 2060 super;

probably by lowering the temperature from lowered power used. something like msi afterburner that limits the max temperature and power used, in firmware.

then they ramped up the power to make the 2070 super.

so if your looking for the new gpu look at the info i gave above and make an informed decision.






RuneSR2
Grand Champion
Unfortunately technical specs don't always translate into real world performance - that's why you need results from a lot of different games to get an idea how different video cards perform in the real world. Also video cards perform differently in games, so you need many games to get reliable results - not just 5 games. You may also need to test different resolutions - because 4K results may not translate well if you use a 2K monitor etc. 
I'd never rely on technical specs for power consumption too - again different cooling systems, different overclocks etc. may cause two 2080 cards to behave quite differently in the real world. 

Oculus Rift CV1, Valve Index & PSVR2, Asus Strix OC RTX™ 3090, i9-10900K (5.3Ghz), 32GB 3200MHz, 16TB SSD
"Ask not what VR can do for you, but what you can do for VR"

hoppingbunny123
Rising Star
the scientific method starts from an observation thats then proven step by step. what i showed was science.

your observation to falsify my observation is just an observation not disproving my proof.


if your observation disagrees with experiment its wrong, and im right.