GTX 1080 vs Overclocked MSI GTX 980 Ti Lightning  

  RSS
Jester
(@jester)
Member Moderator

The results may surprise you. I recently upgraded from an MSI GTX 980 Ti Lightning (thanks to Enad for this) which I ran at 1482 - 1488 MHz core (some were ran at the former, others the latter), stock memory (7100 MHz effective, it couldn't go beyond that) to an MSI GTX 1080 ARMOR 8G OC. Many people suspect this isn't much of an upgrade, but here are some performance comparisons.

The rest of the system is as follows:

  • Windows 10 Pro 64-bit
  • ASUS Maximus VIII Hero Z170
  • Intel Core i7 6700k @ 4.6 GHz
  • G.SKILL 16GB (2 x 8GB) DDR4 3000 CL15
  • Samsung 850 EVO 500GB SSD

Pics:

h0gqzvY.jpg
GTX 980 Ti Lightning left, GTX 1080 Armor right. GTX 980 Ti has far superior packaging, the most luxurious and protective I've ever seen.

xOcBAUd.jpg
GTX 980 Ti Lightning, by far the biggest GPU I've ever handled.

WLGZDLK.jpg

TK0VSv2.jpg
Awful picture of the GTX 1080 Armor, sorry my brother wasn't around with his camera to photograph it and I wasn't about to wait for him.

Synthetic Benchmarks

kgr9Ita.jpg

jCDGRFS.jpg

MSI GTX 980 Ti Lightning (Overclocked)

  • 3DMark Fire Strike: 17457 total, 20774 graphics. Link
  • 3DMark Fire Strike Extreme: 8838 total, 9362 graphics. Link
  • 3DMark Fire Strike Ultra: 4782 total, 4724 graphics. Link
  • 3DMark API Overhead Test: 2,703,609 DX11 multithread, 2,743,922 DX11 single thread, 21,133,382 DX12. Link
  • 3DMark 11 Performance: P21723 total, 27283 graphics. Link
  • 3DMark 11 Extreme: X9064 total, 8600 graphics. Link
  • Unigine Heaven Extreme (Pictured above): 3131 score, 124.3 FPS average, 34.4 FPS minimum, 273.7 FPS maximum
  • Unigine Valley Extreme HD (Pictured above): 4158 score, 99.4 FPS average, 43.2 FPS minimum, 191.6 FPS maximum

MSI GTX 1080 ARMOR 8G OC (Stock)

  • 3DMark Fire Strike: 18562 total, 22768 graphics. Link
  • 3DMark Fire Strike Extreme: 9881 total, 10701 graphics. Link
  • 3DMark Fire Strike Ultra: 5268 total, 5222 graphics. Link
  • 3DMark API Overhead Test: 2,682,433 DX11 multithread, 2,826,200 DX11 single thread, 29,512,677 DX12. Link
  • 3DMark 11 Performance: P23521 total, 31364 graphics. Link
  • 3DMark 11 Extreme: X10172 total, 9794 graphics. Link
  • Unigine Heaven Extreme (Pictured above): 3616 score, 143.6 FPS average, 38.8 FPS minimum, 300.9 FPS maximum
  • Unigine Valley Extreme HD (Pictured above): 4562 score, 109.0 FPS average, 47.5 FPS minimum, 218.1 FPS maximum

MSI GTX 1080 ARMOR 8G OC (overclocked to 11016 MHz effective memory and a typical core clock of 2063 MHz with peaks just over 2100 MHz)

  • 3DMark Fire Strike: 19347 total, 24489 graphics. Link
  • 3DMark Fire Strike Extreme: 10603 total, 11592 graphics. Link
  • 3DMark Fire Strike Ultra: 5617 total, 5587 graphics. Link
  • 3DMark 11 Performance: P24686 total, 34294 graphics. Link
  • 3DMark 11 Extreme: X11081 total, 10761 graphics. Link
  • Unigine Heaven Extreme (Pictured above): 3905 score, 155.0 FPS average, 39.0 FPS minimum, 316.4 FPS maximum
  • Unigine Valley Extreme HD (Pictured above): 4958 score, 118.5 FPS average, 45.7 FPS minimum, 236.7 FPS maximum

I have no comparison for this one since it came out so late, but 3DMark Time Spy: 7555 with 7918 graphics score.

http://www.3dmark.com/spy/64639

Gaming Benchmarks

All of these tests were conducted at 2560 x 1440.

xTXfIb0.jpg

lWFAyYk.jpg

9dqJnLV.jpg

DX11 vs Vulkan and DX12

In these tests I won't be comparing graphics cards, only looking at the overclocked GTX 1080. Check out how it performs in DX12 and Vulkan; Total War: Warhammer for the first, and The Talos Principle for the second. Embarrassing results especially for the latter.

LXQHCOa.jpg
Total War: Warhammer DX11

4BnhTwd.jpg
Total War: Warhammer DX12

Ngz8tTz.jpg

Why so bad? It's because these aren't true DX12 or Vulkan games, rather they were added last minute. CPU usage isn't really any different compared to DX11, I'm pretty sure they do nothing but enable asynchronous shading which NVIDIA can't do at a hardware level. Still, what a joke. By the way here is the log for both of those Talos Principle runs.

Code:
Gfx API: Direct3D11
Duration: 184.0 seconds (18442 frames)
Average: 100.2 FPS (104.4 w/o extremes)
Extremes: 870.4 max, 48.6 min
Sections: AI=6%, physics=1%, sound=1%, scene=75%, shadows=15%, misc=3%
Highs: 1886 in 13.9 seconds (135.7 FPS)
Lows: 2517 in 35.6 seconds (70.8 FPS)
30-60 FPS: 1%
> 60 FPS: 99%
Code:
Gfx API: Vulkan
Duration: 186.3 seconds (11054 frames)
Average: 59.3 FPS (59.7 w/o extremes)
Extremes: 807.4 max, 7.1 min
Sections: AI=3%, physics=0%, sound=1%, scene=90%, shadows=4%, misc=2%
Highs: 32 in 0.4 seconds (80.4 FPS)
Lows: 253 in 5.4 seconds (46.7 FPS)
30-60 FPS: 98%
> 60 FPS: 2%

Sorry I didn't have time for more benchmarking, although I have been doing a lot of actual playing and I can say the difference between the two cards, particularly with the GTX 1080 overclocked, is more than welcome. It was only worthwhile because I sold the GTX 980 Ti Lightning though.

ReplyQuote
Posted : 09/07/2016 5:59 pm
Share:

You Might Also Like

  
Working

Please Login or Register