Disappointed with my 2070 super

[Raidboss]

So i built my first gaming pc a week ago, after a lot of research i came up with a list of parts that seemed ideal for my main goal which is 1440 p gaming. i came up with:

 

Cpu :Ryzen 5 3600 

Gpu: 2070 super

RAM: 16 GB DDR4 3200 MHZ CAS 15

SSD: Wd blue 500 GB

HDD: Seagate 2tb

PSU : EVGA 650W gold+

Cpu cooler: Noctua nh-u9s 

 

So far i tried running Battlefield V (65 fps average), Apex Legends (100+ fps avg). Bannerlords (90 fps avg) . The Witcher 3(85-90 fps avg). All tests were made with settings maxed out on ultra. I am deeply disappointed in the performance of my GPU, before buying it i saw various benchmarks that were showing for instance Battlefield V running at ATLEAST 90 fps average or The Witcher 3 at ATLEAST 110 fps average. What could be causing this huge drop in performance? I have been monitoring temps and i dont really think throttling could be the issue since after a long session my GPU on average sits at around 75 ° (167 F), which is not great but not awful considering its pretty warm where i live, the highest i have seen my GPU go is 78°(172F), exactly the same for the highest temp of my CPU, which in turn on average sits at around 70°(158 F) after a long session. Any ideas? I kinda feel scammed by the benchmarks not gonna lie

 

[Mateyyy]

Either you've got your render scale set to 200% or you've got DXR turned on, because BF5 ran at 100+ FPS almost all the time with my prior 2070 non-Super.

[5x5]

Check that you've not enabled Hairworks, DXR, DLSS or any other supersampling setting

[Raidboss]

I do have DXR in BF5, it's set by default and for some reason i cant turn it off.. but if that's the problem with BF5, what about the other games like the Witcher? There's no Ray tracing there as far as i know

 

[Raidboss]

I had Haiworks on in The Witcher. tried turning it off and i barely got a 3-4 fps boost, i dont think thats the problem there. By the way, wont turning off all supersampling give a worse image quality? Dont benchmarkers test the games with some sort of supersampling?

[Sorenson]

A few recommendations for your in game settings: turn nvidia stuff off (dxr, hairworks, etc), turn shadows and lighting to high, turn anti aliasing to medium/high, turn post processing to high, and turn down reflections to high. I can almost guarantee you won’t notice a difference and will get much better performance. Ultra settings with high antialiasing are fps killers with little benefit. 

[Raidboss]

2 minutes ago, Sorenson said:

A few recommendations for your in game settings: turn nvidia stuff off (dxr, hairworks, etc), turn shadows and lighting to high, turn anti aliasing to medium/high, turn post processing to high, and turn down reflections to high. I can almost guarantee you won’t notice a difference and will get much better performance. Ultra settings with high antialiasing are fps killers with little benefit. 

Thanks for all the suggestions, but my point is that whan i bought the 2070 super i thought i bought a card capable of giving me a great 1440p gaming experience with everything maxed out, as seen in benchmarks, apparently to get the same results i need to turn down a lot of stuff? That's truly a case of false advertisement

[Gaires]

25 minutes ago, Raidboss said:

Thanks for all the suggestions, but my point is that whan i bought the 2070 super i thought i bought a card capable of giving me a great 1440p gaming experience with everything maxed out, as seen in benchmarks, apparently to get the same results i need to turn down a lot of stuff? That's truly a case of false advertisement

Did the benchmarks specify what settings they were using, because a lot of them turn down or even turn off some settings and still say they're testing "ultra settings"?

Or they might have used faster ram, better ssd, their cooling might have been better, they might use frest install of windows with no programs running on the background etc.

There are a lot of variables in play here, and reviews should be used as a reference, and not as an absolute truth.

[Gohardgrandpa]

You went into it expecting way too much. I have a rx5700xt and I can’t max everything out @1080p 144fps. Some games run better then others and you just need to find out what settings up the framerate without visual making your game look like a mashed potato 

[Mateyyy]

1 hour ago, Raidboss said:

By the way, wont turning off all supersampling give a worse image quality? Dont benchmarkers test the games with some sort of supersampling?

Supersampling is the most demanding form of anti-aliasing, as it basically renders the game at a higher resolution. After a point, the "better image quality" is pretty much just a placebo, unless you often take screenshots of your game and zoom in 300%.

At 1440p you could just use TAA even and the image would still be sharp in most cases, otherwise you could also add a sharpening filter in Nvidia Freestyle.

57 minutes ago, Raidboss said:

Thanks for all the suggestions, but my point is that whan i bought the 2070 super i thought i bought a card capable of giving me a great 1440p gaming experience with everything maxed out, as seen in benchmarks, apparently to get the same results i need to turn down a lot of stuff? That's truly a case of false advertisement

No, that's just having false expectations. Also, most "maxed out" benchmarks don't include maxing out settings like supersampling, which again, don't offer a great enough visual improvement to justify the massive hit in performance, hence why most people don't use such settings.

[Raidboss]

1 hour ago, Gaires said:

Did the benchmarks specify what settings they were using, because a lot of them turn down or even turn off some settings and still say they're testing "ultra settings"?

Or they might have used faster ram, better ssd, their cooling might have been better, they might use frest install of windows with no programs running on the background etc.

There are a lot of variables in play here, and reviews should be used as a reference, and not as an absolute truth.

Sometimes, they specified which settings they were turned off. But even then, if i turn off their same exact settings i won't get near their performance. that's what bothers me

 

1 hour ago, Gohardgrandpa said:

You went into it expecting way too much. I have a rx5700xt and I can’t max everything out @1080p 144fps. Some games run better then others and you just need to find out what settings up the framerate without visual making your game look like a mashed potato 

I thought those tests could be somewhat reliable, i dont expect a benchmark to show me a game running at 110 fps and then having to play it at 80! A small difference should be acceptable of course but not in this order of magnitude

1 hour ago, Mateyyy said:

Supersampling is the most demanding form of anti-aliasing, as it basically renders the game at a higher resolution. After a point, the "better image quality" is pretty much just a placebo, unless you often take screenshots of your game and zoom in 300%.

At 1440p you could just use TAA even and the image would still be sharp in most cases, otherwise you could also add a sharpening filter in Nvidia Freestyle.

No, that's just having false expectations. Also, most "maxed out" benchmarks don't include maxing out settings like supersampling, which again, don't offer a great enough visual improvement to justify the massive hit in performance, hence why most people don't use such settings.

I dont know why you're telling me that i have false expectations. Buying a 2070 super and expecting it to run 4k 100 fps is false expectation. I bought a 2070 super expecting a certain quality thast has not been met. I'm not asking to get 144 fps on ultra on AAA titles, im just asking something better than 65-70 fps on BF5, a game that came out 2 years ago. 

[Mateyyy]

5 minutes ago, Raidboss said:

I dont know why you're telling me that i have false expectations. Buying a 2070 super and expecting it to run 4k 100 fps is false expectation. I bought a 2070 super expecting a certain quality thast has not been met. I'm not asking to get 144 fps on ultra on AAA titles, im just asking something better than 65-70 fps on BF5, a game that came out 2 years ago. 

Well, expecting to run BF5 at more than 70FPS at 1440p Ultra with DXR is a false expectation. It's well known how demanding it is on all current-gen RTX cards, even the 2080Ti.

Just turn off DXR. I'm pretty certain you will notice 0 difference in visual quality, in such a fast paced game where the only thing that actually benefits from raytracing is reflections, which are a minor detail considering the game's maps and general style.

[Dravinianv2]

The benchmarks I saw with Ultra 1440p 103fps in Battlefield V had a i9-9900k clocked at 5ghz with 16gb 3400mhz RAM.

 

Maybe that is the difference?

[Sorenson]

2 hours ago, Raidboss said:

Thanks for all the suggestions, but my point is that whan i bought the 2070 super i thought i bought a card capable of giving me a great 1440p gaming experience with everything maxed out, as seen in benchmarks, apparently to get the same results i need to turn down a lot of stuff? That's truly a case of false advertisement

You’re getting 90+ FPS with everything cranked at 1440p with antialiasing and DXR enabled (in some games). That is a great experience. If you want ~120+ FPS you need to turn down settings. You would need to do this even on a 2080ti. I’m sorry it’s not what you were expecting, but ultra settings are resource wasters. Try turning down the settings I mentioned to high and DXR to medium and let me know if you see a quality difference. 
 

I had a 2080 super in my last computer and I still had to turn down settings to get 100+ FPS. In my new build I’m using a 2070 super because it’s close enough and turning down settings gives almost the exact same experience. 

[Raidboss]

2 hours ago, Mateyyy said:

Well, expecting to run BF5 at more than 70FPS at 1440p Ultra with DXR is a false expectation. It's well known how demanding it is on all current-gen RTX cards, even the 2080Ti.

Just turn off DXR. I'm pretty certain you will notice 0 difference in visual quality, in such a fast paced game where the only thing that actually benefits from raytracing is reflections, which are a minor detail considering the game's maps and general style.

I’ll try turning it off, game looked so nice im afraid of touching the settings. 
 

 

2 hours ago, Dravinianv2 said:

The benchmarks I saw with Ultra 1440p 103fps in Battlefield V had a i9-9900k clocked at 5ghz with 16gb 3400mhz RAM.

 

Maybe that is the difference?

it may be, but as far as i know my Cpu shouldn’t really bottleneck my Gpu at all, but i could be wrong. I can tell you my Cpu isnt really near full load when gaming anyway

2 hours ago, Sorenson said:

You’re getting 90+ FPS with everything cranked at 1440p with antialiasing and DXR enabled (in some games). That is a great experience. If you want ~120+ FPS you need to turn down settings. You would need to do this even on a 2080ti. I’m sorry it’s not what you were expecting, but ultra settings are resource wasters. Try turning down the settings I mentioned to high and DXR to medium and let me know if you see a quality difference. 
 

I had a 2080 super in my last computer and I still had to turn down settings to get 100+ FPS. In my new build I’m using a 2070 super because it’s close enough and turning down settings gives almost the exact same experience. 

i will try and follow your suggestions. It boggles me that even a 2080 super is not enough, Maybe 1440p is still a bit too demanding on GPUs at this stage

[Dravinianv2]

There is a video here:

https://www.youtube.com/watch?v=zYVokmuchSE

 

Where he benchmarks a number of games with your set up.

 

Games differ but sometimes he is getting 60 fps (Ghost Recon) for example.

 

Might be worth a watch, at least you may be able to put to rest whether you are not getting what you should be getting as he appears to show the setings....so you can see whether you are matching him.

[Raidboss]

14 hours ago, Dravinianv2 said:

There is a video here:

https://www.youtube.com/watch?v=zYVokmuchSE

 

Where he benchmarks a number of games with your set up.

 

Games differ but sometimes he is getting 60 fps (Ghost Recon) for example.

 

Might be worth a watch, at least you may be able to put to rest whether you are not getting what you should be getting as he appears to show the setings....so you can see whether you are matching him.

These results seem more plausible, thanks. Should have watched this video before the others

[Zberg]

It appears you were expecting too much from it.  The 2080ti cant get crazy high rates at max settings on every game at 1440p, and that card has significantly more power than a 2070S.  I have a 165 hz refresh monitor, guarantee I wouldn't hit it in any of the titles you mention (except apex legends, but even then its not guaranteed, and has dips in the 100-120's).

 

Everything maxed (DXR, hairworks, ultra/extreme settings) at 1440p is a huge ask graphically.  I have 2 games specifically where I get dips to 50's/60's with all those settings turned on at 1440p.   Frankly I am surprised a 2070S is getting rates as good as you were.

 

Sweet spot is usually going down 1 setting (extreme to ultra, or ultra to high for example), turning DXR to low (if you like it) or off (if you dont), and hairworks off.  Rarely will doing those things be that noticable, and you usually get a big performance bump from doing them

 

If you see benchmark rundowns from channels like hardware unboxed and the like, you will often see in the settings in the top corner of the screen that they use settings similar to what I have above.  At the minimum they usually turn hairworks off, DXR to low or off.

[VerTTi]

On 4/18/2020 at 7:59 PM, Raidboss said:

So i built my first gaming pc a week ago, after a lot of research i came up with a list of parts that seemed ideal for my main goal which is 1440 p gaming. i came up with:

 

Cpu :Ryzen 5 3600 

Gpu: 2070 super

RAM: 16 GB DDR4 3200 MHZ CAS 15

SSD: Wd blue 500 GB

HDD: Seagate 2tb

PSU : EVGA 650W gold+

Cpu cooler: Noctua nh-u9s 

 

So far i tried running Battlefield V (65 fps average), Apex Legends (100+ fps avg). Bannerlords (90 fps avg) . The Witcher 3(85-90 fps avg). All tests were made with settings maxed out on ultra. I am deeply disappointed in the performance of my GPU, before buying it i saw various benchmarks that were showing for instance Battlefield V running at ATLEAST 90 fps average or The Witcher 3 at ATLEAST 110 fps average. What could be causing this huge drop in performance? I have been monitoring temps and i dont really think throttling could be the issue since after a long session my GPU on average sits at around 75 ° (167 F), which is not great but not awful considering its pretty warm where i live, the highest i have seen my GPU go is 78°(172F), exactly the same for the highest temp of my CPU, which in turn on average sits at around 70°(158 F) after a long session. Any ideas? I kinda feel scammed by the benchmarks not gonna lie

 

My RTX 2070 SUPER paired with R5 3600 runs 90+ fps 1440p on BFV multiplayer with everything maxed out so yeah there is something wrong. Did you perform a clean install of drivers?

[Zberg]

33 minutes ago, VerTTi said:

My RTX 2070 SUPER paired with R5 3600 runs 90+ fps 1440p on BFV multiplayer with everything maxed out so yeah there is something wrong. Did you perform a clean install of drivers?

 

 

Yours may be a more OC'd card or a better one?  The above video shows it pretty much sitting in the 70's, dipping to the 60's with RTX on.

 

Hardware unboxed also found that they got their 2070 Super to have framerates such as you are saying on ultra quality 1440p...with RTX off.

 

It seems the numbers you are getting are an exception and you might have just gotten a baller GPU.

 

 

[Dravinianv2]

Or the way he is measuring FPS is inaccurate.

[VerTTi]

Well i wont vouch for origins fps counter but im getting on avrg 90 fps with RTX on and all settings max. And yeah the card is OCD but still without it would pull way more frames than OPs (just tested 79 fps on avrg with stock)

Btw the first map is foliage galore hence the lower frames and yeah we all know snow (i avrg 80fps on those maps) the maps ive been playing are the new pasific ones.

[Dravinianv2]

21 minutes ago, VerTTi said:

Well i wont vouch for origins fps counter but im getting on avrg 90 fps with RTX on and all settings max. And yeah the card is OCD but still without it would pull way more frames than OPs (just tested 79 fps on avrg with stock)

Btw the first map is foliage galore hence the lower frames and yeah we all know snow (i avrg 80fps on those maps) the maps ive been playing are the new pasific ones.

Well there is this video:

Seems to show 80-100 fps too.

[Dravinian]

But bear in mind in that video he has DXR off, and DX12 off.

 

So no ray tracing.  So getting high FPS isn't really a problem without those, it is with those that hte problems arise.

 

Is that what you are saying VerTTi, that you are getting 100fps without ray tracing?  I am not sure what the OP is getting without ray tracing, but I think his complaint was that with ray tracing he wasn't getting 100 fps/