3070 FE, not what I expected??

[mikestrong88]

Hey everyone, 

 

Got my 3070 FE delivered today and I am confused. Having looked at various benchmarks, fps counters and general reviews, I don't seem to be getting the numbers that others are and I have no clue why. Running 1440p with the following;

 

R5 3600

x570 TUF MB

32GB 3600mhz Ram

O11D case (temps on GPU never exceed 72c)

No riser cable being used

 

This link for example, shows ultra settings with the 3070 on COD Warzone, with the R5 3600 and he is averaging 120+, where as mine is closer to 100fps. 

 

This link, goes into detail with the 10900k but also shows the varying CPUs and at the bottom & shows the R5 3600 and says it should be around 150 fps, but even in MP I am only getting around 120fps

 

I cant work out why mine would be up to 20% less than these links? It not even like they're Nvidia marketing specs that I would expect to be exaggerated. They seem to be legit and running the same hardware I am so I can't see why there would be any bottlenecks for the other hardware I am using. I also ran a few heaven benchmarks and seemed to be getting around 50pts less than some of the reviews I had seen. For reference, GPU never exceeds 72c and CPU around 51C

 

Thanks 

[jaslion]

Warzone is just a badly optimized game.

[mikestrong88]

Just now, jaslion said:

Warzone is just a badly optimized game.

I get that, but seeing other people getting these results in that badly optimised game makes me think there's something more to my low figures. Same in MP too 

[jaslion]

Just now, mikestrong88 said:

I get that, but seeing other people getting these results in that badly optimised game makes me think there's something more to my low figures. Same in MP too 

I've seen most people cap out at 120fps with a 3600 so I'm pretty sure that is your fps cap basically. Got anything else running in the background?

[mikestrong88]

Just now, jaslion said:

I've seen most people cap out at 120fps with a 3600 so I'm pretty sure that is your fps cap basically. Got anything else running in the background?

Not on the last test no, figured that may be the issue but nothing changed when I closed discord and chrome

[Senzelian]

Best thing is to compare repeatable benchmarks and not in game scenes. That way you can make sure that your card performs as it should.

[mikestrong88]

1 minute ago, Senzelian said:

Best thing is to compare repeatable benchmarks and not in game scenes. That way you can make sure that your card performs as it should.

For example heaven? I ran a few of those and again was around 150pts under the scores that someone else got. In honesty though they had a much better cpu than me so idk if that would affect that specific benchmark 

[Cyracus]

12 minutes ago, mikestrong88 said:

This link for example, shows ultra settings with the 3070 on COD Warzone, with the R5 3600 and he is averaging 120+, where as mine is closer to 100fps. 

Their 3600 is overclocked to 4.4ghz

[mikestrong88]

1 minute ago, Cyracus said:

Their 3600 is overclocked to 4.4ghz

I missed that! Good spot. Maybe that's where I'm missing out. Doesn't explain the other one but that's a great spot. 

[Falkentyne]

14 minutes ago, mikestrong88 said:

Hey everyone, 

 

Got my 3070 FE delivered today and I am confused. Having looked at various benchmarks, fps counters and general reviews, I don't seem to be getting the numbers that others are and I have no clue why. Running 1440p with the following;

 

R5 3600

x570 TUF MB

32GB 3600mhz Ram

O11D case (temps on GPU never exceed 72c)

No riser cable being used

 

This link for example, shows ultra settings with the 3070 on COD Warzone, with the R5 3600 and he is averaging 120+, where as mine is closer to 100fps. 

 

This link, goes into detail with the 10900k but also shows the varying CPUs and at the bottom & shows the R5 3600 and says it should be around 150 fps, but even in MP I am only getting around 120fps

 

I cant work out why mine would be up to 20% less than these links? It not even like they're Nvidia marketing specs that I would expect to be exaggerated. They seem to be legit and running the same hardware I am so I can't see why there would be any bottlenecks for the other hardware I am using. I also ran a few heaven benchmarks and seemed to be getting around 50pts less than some of the reviews I had seen. For reference, GPU never exceeds 72c and CPU around 51C

 

Thanks 

What speed is your CPU at ?  On a R5 3600, that matters.

Also he had DXR disabled in his 1440p test on your first video link.

[DoctorNick]

4 minutes ago, mikestrong88 said:

For example heaven? I ran a few of those and again was around 150pts under the scores that someone else got. In honesty though they had a much better cpu than me so idk if that would affect that specific benchmark 

What resolution you run in WZ? Edit: oh 1440p. Try Superpostion instead of heaven and post your score

[Mark Kaine]

6 minutes ago, Senzelian said:

Best thing is to compare repeatable benchmarks and not in game scenes. That way you can make sure that your card performs as it should.

Also a lot of videos on yt are simply fake... You can't know anything they present is true or not. Much better is something like Firestrike where you can compare to a plethora of same hardware results for example. 

 

 

I'm pretty sure there's cheaters too, but it's much harder than on yt which is basically free for all, yt doesn't care, they only care about clicks. 

[mikestrong88]

2 minutes ago, Falkentyne said:

What speed is your CPU at ?  On a R5 3600, that matters.

Also he had DXR disabled in his 1440p test on your first video link.

Stock with PBO so turbos to 4.2ghz from memory. I'd have to check tomorrow but I'm 99% sure. Yeah I have no Raytracing on either 

[mikestrong88]

3 minutes ago, DoctorNick said:

What resolution you run in WZ? Edit: oh 1440p. Try Superpostion instead of heaven and post your score

I'll do that tomorrow and post results  

[mikestrong88]

1 minute ago, Mark Kaine said:

Also a lot of videos on yt are simply fake... You can't know anything they present is true or not. Much better is something like Firestrike where you can compare to a plethora of same hardware results for example. 

 

 

I'm pretty sure there's cheaters too, but it's much harder than on yt which is basically free for all, yt doesn't care, they only care about clicks. 

So this potentially could be a bullshit result? 

 

The other site seems fairly credible and I'm waaaay off their scores 

[Cyracus]

4 minutes ago, mikestrong88 said:

I missed that! Good spot. Maybe that's where I'm missing out. Doesn't explain the other one but that's a great spot. 

Many factors from ram speed to cpu cooling and overclocks that could affect their numbers.
Unfortunately they don't appear to include any other information about the test rig gpucheck drew those numbers from so it's impossible to know where that discrepancy stems

[Mark Kaine]

6 minutes ago, mikestrong88 said:

So this potentially could be a bullshit result? 

 

The other site seems fairly credible and I'm waaaay off their scores 

Well of course it could be... There's no way to tell without actually fact checking. And even if it's not intentionally they could have some OCs and whatnot... 

 

Really the best thing to do is Firestrike and Timespy, you get to compare to "same hardware" and you can also see if they OC or not if you check frequencies and stuff. 

 

It's a really good test suit to find out if your hardware works as intended. You prob won't even finish Firestrike if there are hw/sw issues, the test knows no mercy lol

[Senzelian]

12 minutes ago, mikestrong88 said:

For example heaven? I ran a few of those and again was around 150pts under the scores that someone else got. In honesty though they had a much better cpu than me so idk if that would affect that specific benchmark 

3D Mark FireStrike is usually the best starting point. They have individual graphics scores, which take the CPU mostly out of the equation and you can compare results very easily.

[mikestrong88]

2 minutes ago, Mark Kaine said:

Well of course it could be... There's no way to tell without actually fact checking. And even if it's not intentionally they could have some OCs and whatnot... 

 

Really the best thing to do is Firestrike and Timespy, you get to compare to "same hardware" and you can also see if they OC or not if you check frequencies and stuff. 

 

It's a really good test suit to find out if your hardware works as intended. You prob won't even finish Firestrike if there are hw/sw issues, the test knows no mercy lol

Sweet, thanks! I'll do that tomorrow and see how I compare. A few % doesn't concern me but if its like 20% as I thought here, it's kinda crappy. 

 

A bad feeling when you've just got this new GPU and you're a little underwhelmed. 

[DoctorNick]

8 minutes ago, mikestrong88 said:

I'll do that tomorrow and post results  

Okay here is something for comparison, sig rig:

1080 ti@stock 1987-1850mhz boost: