- Joined
- Feb 22, 2009
- Messages
- 762 (0.14/day)
| System Name | Lenovo 17IMH05H |
|---|---|
| Processor | Core i7 10750H |
| Video Card(s) | GTX 1660 Ti |
| Audio Device(s) | SSL2 |
| Software | Windows 10 Pro 22H2 |
| Benchmark Scores | i've got a shitload of them in 15 years of TPU membership |
Following my ''research'', in which i sorta have proven that NVIDIA does not gimp it's previous generation video cards with new driver updates, i have decided to do one final showdown while i still have GTX 780 and my friend has GTX 970 - to conclude the position of these VRAM lacklusting video cards in modern gaming titles. In the years when these cards were released (2013 & 2014) both of them could drive through every game at that time on 1920x1080 resolution with max settings no problem, sorta. Let's see one more time what can these cards do in the games released from 2015 till 2018. Only Crysis 3 and and Company of Heroes 2 will be revisited from 2013, just because they are still really demanding titles, and there have not been sequels released yet.
TEST SETUP
Intel Core i7 4770K 4500 MHz OC
Gigabyte GA-Z97X-Gaming 3
Crucial Ballistix Tactical 32 GB DDR3 1600 MHz CL8
Samsung 850 Evo M.2 500 GB
Western Digital SE WD2000F9YZ 2 TB
Zotac GeForce GTX 970 4 GB {core overclocked to 1245 MHz/memory overclocked to 1950 MHz}
Zotac GeForce GTX 780 3 GB {core overclocked to 1045 MHz/memory overclocked to 1750 MHz}
--------------------------------------------------------------------------------------------------
You can see how much these cards have been overclocked from their factory settings in GPU-Z. Higher values would crash random games, so i had to settle for what you see. Still this is higher than any factory OC you can find. The best GTX 780 is probably KFA2 HOF version with 1003 MHz core clock out of the box, while the best GTX 970 is probably Colorful iGame version with 1200 MHz core clock out of the box.
This might be the last test i perform this year for TPU as well as the last one made on Windows 7.
In all you will be looking at 25 games tested at best playable settings selected - meaning that if the card can not run the game maxed out, the settings are lowered for playable experience. For first person shooter games and racing games this means that at least 50 FPS on average must be achieved on both cards, while for strategy games this means that at least 30 FPS as minimum be achieved. To resolute these achievements overclocking both cards was essential, as you will later find out.
LET'S START!
ASHES OF SINGULARITY (2016)
This most CPU intense game is equally just for both cards when run on high settings with no more than 2X MSAA.
BATTLEFIELD 1 (2016)
With T-AA the game is smooth on both cards.
CALL of DUTY WW2 (2017)
GTX 780 seems to have no trouble even on max settings.
COMPANY of HEROES 2 (2013)
Preferably AA should be set at medium for best smooth gameplay, but even with AA high selected both cards are equally capable.
CRYSIS 3 (2013)
By now this is the oldest title in my test book, which still looks amazing. Enabling SMAA to 4X severely punishes both cards. Was not surprised to find GTX 780 faster.
DEUS EX MANKIND DIVIDED (2016)
This benchmark shows the worst case scenario for GTX 780, because in typical levels (Desperate Measures, System Rift, Criminal Past) the performance is way better.
DIRT 4 (2017)
You will find out that GTX 780 can match GTX 970 in many modern titles, not just in old ones.
DOOM (2016)
Out of all the tested ames, Doom shows the biggest difference between Kepler and Maxwell possible. This is the only game in which GTX 970 G1 managed to beat GTX 780 Ti GHz Edition in all of my previous tests ever made.
F1 2017 (2017)
Nothing to say here.
FAR CRY 5 (2018)
The newest Far Cry can be maxed out with GTX 780 no trouble, although for best experience one should slightly lower the settings to get 60 FPS average instead of 53...
FORSTPUNK (2018)
Though the game is limited to 60 FPS, both cards have slight trouble maintaining that number, so for perfect 60 FPS v-sync some settings should be adjusted minorly.
GHOST RECON WILDLANDS (2017)
If you want to enjoy this game with a GTX 780, high settings with T-AA should be the max you aim for.
HELLBLADE SENUA'S SACRIFICE (2017)
Not sure about the AA used in this game, but both cards can easy max the game out.
MASS EFFECT ANDROMEDA (2017)
Settings should be drooped from max to high and T-AA should be selected for GTX 780 to feel smooth.
MIDDLE EARTH SHADOW of WAR (2017)
The official benchmark was corrupted, so i had to start the game the common way to see how things look at very high preset and it seems GTX 780 is faster in this game.
PREY (2017)
I feel like there could be more AA modes in this game to punish the cards more.
PROJECT CARS 2 (2017)
I find it bizarre that the sequel has way less graphical options and AA settings than the first game. Whatever.
RISE OF TOMB RAIDER (2015)
The second benchmark was tested out of the 3. Actually both cards can max out this game even at higher settings than high, but definitely not with SSAA modes.
SNIPER ELITE 4 (2017)
TITANFALL 2 (2016)
No wonder GTX 780 is faster than GTX 970 on a source engine game.
TOTAL WAR WARHAMMER 2 (2017)
But i do wonder howcome GTX 780 is the faster card in this game, and by quite a margin. The official benchmark of this game is quite stable, each time providing the results with just 0,5 % difference.
VANISHING of ETHAN CARTER REDUX (2015)
Results are very similar as in Prey.
WATCH DOGS 2 (2016)
For best experience one should slightly lower the settings even more on both cards, but still 50 FPS average is possible even on very high preset. Enabling T-AA instead of SMAA greatly increases performance.
WITCHER 3 WILD HUNT (2015)
50 FPS is possible even on max settings, but you will definitely want to lower the settings to high for that really smooth 60 FPS. Also, GTX 970 is typically faster in other scenes, but is equal at the start. Random luck.
WRECKFEST (2018)
By now i have noticed that GTX 780 and GTX 970 can max out any racing game i throw even with 4X MSAA!
------------------------------------------------------------------------------------------------------------------------------------
While this is not about GTX 780 vs. GTX 970, by calculating the average performance out of the 25 tested games, GTX 970 at it's best stable oc was just 8% faster than GTX 780 at it's best stable oc. Now this can vary alot within the given situation. GTX 780 manages to win in Crysis 3, Shadow of War, Titanfall 2 and Warhammer 2, with many ties too, yet GTX 970 has a hard-to-believe 40 % lead in Doom alone... According to the original Techpowerup GTX 970 SLI review stock GTX 970 was 14 % faster than GTX 780 at 1080p, which is fine by me, as i never reference anything stock anyway. But the fact of the mater is that both cards are still GREAT for modern games and can absolutely max out many of them. Second, there was not a single game which needed medium settings - even the toughest ones, which could not be played on max settings could still be played on high presets.
So what do you think about your GTX 780 and GTX 970? How much long 3 ~ 3.5 GB VRAM will last you!!!?
TEST SETUP
Intel Core i7 4770K 4500 MHz OC
Gigabyte GA-Z97X-Gaming 3
Crucial Ballistix Tactical 32 GB DDR3 1600 MHz CL8
Samsung 850 Evo M.2 500 GB
Western Digital SE WD2000F9YZ 2 TB
Zotac GeForce GTX 970 4 GB {core overclocked to 1245 MHz/memory overclocked to 1950 MHz}
Zotac GeForce GTX 780 3 GB {core overclocked to 1045 MHz/memory overclocked to 1750 MHz}
--------------------------------------------------------------------------------------------------
You can see how much these cards have been overclocked from their factory settings in GPU-Z. Higher values would crash random games, so i had to settle for what you see. Still this is higher than any factory OC you can find. The best GTX 780 is probably KFA2 HOF version with 1003 MHz core clock out of the box, while the best GTX 970 is probably Colorful iGame version with 1200 MHz core clock out of the box.
This might be the last test i perform this year for TPU as well as the last one made on Windows 7.
In all you will be looking at 25 games tested at best playable settings selected - meaning that if the card can not run the game maxed out, the settings are lowered for playable experience. For first person shooter games and racing games this means that at least 50 FPS on average must be achieved on both cards, while for strategy games this means that at least 30 FPS as minimum be achieved. To resolute these achievements overclocking both cards was essential, as you will later find out.
LET'S START!
ASHES OF SINGULARITY (2016)
This most CPU intense game is equally just for both cards when run on high settings with no more than 2X MSAA.
BATTLEFIELD 1 (2016)
With T-AA the game is smooth on both cards.
CALL of DUTY WW2 (2017)
GTX 780 seems to have no trouble even on max settings.
COMPANY of HEROES 2 (2013)
Preferably AA should be set at medium for best smooth gameplay, but even with AA high selected both cards are equally capable.
CRYSIS 3 (2013)
By now this is the oldest title in my test book, which still looks amazing. Enabling SMAA to 4X severely punishes both cards. Was not surprised to find GTX 780 faster.
DEUS EX MANKIND DIVIDED (2016)
This benchmark shows the worst case scenario for GTX 780, because in typical levels (Desperate Measures, System Rift, Criminal Past) the performance is way better.
DIRT 4 (2017)
You will find out that GTX 780 can match GTX 970 in many modern titles, not just in old ones.
DOOM (2016)
Out of all the tested ames, Doom shows the biggest difference between Kepler and Maxwell possible. This is the only game in which GTX 970 G1 managed to beat GTX 780 Ti GHz Edition in all of my previous tests ever made.
F1 2017 (2017)
Nothing to say here.
FAR CRY 5 (2018)
The newest Far Cry can be maxed out with GTX 780 no trouble, although for best experience one should slightly lower the settings to get 60 FPS average instead of 53...
FORSTPUNK (2018)
Though the game is limited to 60 FPS, both cards have slight trouble maintaining that number, so for perfect 60 FPS v-sync some settings should be adjusted minorly.
GHOST RECON WILDLANDS (2017)
If you want to enjoy this game with a GTX 780, high settings with T-AA should be the max you aim for.
HELLBLADE SENUA'S SACRIFICE (2017)
Not sure about the AA used in this game, but both cards can easy max the game out.
MASS EFFECT ANDROMEDA (2017)
Settings should be drooped from max to high and T-AA should be selected for GTX 780 to feel smooth.
MIDDLE EARTH SHADOW of WAR (2017)
The official benchmark was corrupted, so i had to start the game the common way to see how things look at very high preset and it seems GTX 780 is faster in this game.
PREY (2017)
I feel like there could be more AA modes in this game to punish the cards more.
PROJECT CARS 2 (2017)
I find it bizarre that the sequel has way less graphical options and AA settings than the first game. Whatever.
RISE OF TOMB RAIDER (2015)
The second benchmark was tested out of the 3. Actually both cards can max out this game even at higher settings than high, but definitely not with SSAA modes.
SNIPER ELITE 4 (2017)
TITANFALL 2 (2016)
No wonder GTX 780 is faster than GTX 970 on a source engine game.
TOTAL WAR WARHAMMER 2 (2017)
But i do wonder howcome GTX 780 is the faster card in this game, and by quite a margin. The official benchmark of this game is quite stable, each time providing the results with just 0,5 % difference.
VANISHING of ETHAN CARTER REDUX (2015)
Results are very similar as in Prey.
WATCH DOGS 2 (2016)
For best experience one should slightly lower the settings even more on both cards, but still 50 FPS average is possible even on very high preset. Enabling T-AA instead of SMAA greatly increases performance.
WITCHER 3 WILD HUNT (2015)
50 FPS is possible even on max settings, but you will definitely want to lower the settings to high for that really smooth 60 FPS. Also, GTX 970 is typically faster in other scenes, but is equal at the start. Random luck.
WRECKFEST (2018)
By now i have noticed that GTX 780 and GTX 970 can max out any racing game i throw even with 4X MSAA!
------------------------------------------------------------------------------------------------------------------------------------
While this is not about GTX 780 vs. GTX 970, by calculating the average performance out of the 25 tested games, GTX 970 at it's best stable oc was just 8% faster than GTX 780 at it's best stable oc. Now this can vary alot within the given situation. GTX 780 manages to win in Crysis 3, Shadow of War, Titanfall 2 and Warhammer 2, with many ties too, yet GTX 970 has a hard-to-believe 40 % lead in Doom alone... According to the original Techpowerup GTX 970 SLI review stock GTX 970 was 14 % faster than GTX 780 at 1080p, which is fine by me, as i never reference anything stock anyway. But the fact of the mater is that both cards are still GREAT for modern games and can absolutely max out many of them. Second, there was not a single game which needed medium settings - even the toughest ones, which could not be played on max settings could still be played on high presets.
So what do you think about your GTX 780 and GTX 970? How much long 3 ~ 3.5 GB VRAM will last you!!!?
Last edited:
. This ended this month, however, as i finally went to Windows 10. I will be also adding frame times too next year. I also really want to do some kind of funny retro benchmark comparing first generation DX10 cards some time in the future, the kind of thing i started my GPU benchmarks in the first place, in oc.uk forums. Plenty of time to start ''collecting'' old video cards, he he.
