I ran Heaven Benchmark using the new EVGA RTX 3070 FTW3. Comparing my results to those on the leaderboard, it's pretty disappointing considering it falls in line with results of 1080's and below. Anyone have any thoughts about the results? Do they seem low? What kind of results are you getting with your card with these settings?
 
FPS: 167.0
Score: 4206
Min FPS: 9.4
Max FPS: 345.5

I was running the settings I saw in a video from JaystwoCents.


I ran the test again and got slightly higher Min FPS and a slightly overal score but still in the gtx 1080 range.

Also, I haven't been wowed by the cards game performance either. I play a few esports games Quake Champions and America's Army which saw no increase in FPS. I have also been testing in Pubg and I see no improvement over my 1080.

I have read a few forums posts on another forum where people are saying they haven't seen any improvements (3070 or 3080) over whatever they replaced. Troubleshooting, I have uninstalled the gpu drivers a few times using DDU in safe mode. I'm running the latest nvidia drivers 457.30. I don't know if there is a problem, I was just expecting more with this card.
 
System Specs
i7 8700k | MSI Z370 GAMING PRO CARBON | 32GB G.SKILL | EVGA RTX 3070 FTW3 | Samsung 970 EVO Plus 1TB | Seagate 4TB HDD | EVGA SuperNOVA 850 G3 | Corsair Hydro Series H100i V2 Extreme Performance | Windows 10 Pro | Alienware 240hz | Phanteks Eclipse P400S

5740 score in 3080FTW3  My lowest FPS was 11.5  seems even the 7900X can't keep the card feed fast enough.
 
Also what's the URL for the leader board page you are referencing?  At the page of the software maker they do not seem to have a Heaven leader board.

scott@bjorn3d
5740 score in 3080FTW3  My lowest FPS was 11.5  seems even the 7900X can't keep the card feed fast enough.
 
Also what's the URL for the leader board page you are referencing?  At the page of the software maker they do not seem to have a Heaven leader board.

I only saw the leaderboard after creating an account. I haven't figured out how to upload results though.

Etyrnus
Is your monitor 1080p, 1440, or 4k?

Alienware 240hz 1080p

Etyrnus
Honestly, you're probably CPU bound.  

Yup, this.

After a certain amount of posts you will be able to put links and pics on your response

arestavo
It takes 10 posts to be able to post links.

Also, if GPU intensive benchmarks like 3DMark Time Spy, Time Spy Extreme, and Port Royal have high GPU usage (they're all rendered at 1440P or higher), and most of your games have low GPU usage - yeah, CPU bound.

I believe Superposition is free benchmark, and has the ability to change resolutions. I don’t believe 1080P maxes out GPU usage, but it should be much higher than 30 to 40%.

A game that DOES scale fairly well even at 1080P is Doom Eternal. If that game exhibits high GPU usage, then you have your answer again.

When you say CPU bound, you mean limited by the cpu? If that's what you mean, I'm not sure how that can be if in game the cpu is not being pushed very hard. I'll test more games tonight and post some results. Also, I was using Heaven as the benchmark tool, is that not a good one to use?
 

TMoseley2

arestavo
It takes 10 posts to be able to post links.

Also, if GPU intensive benchmarks like 3DMark Time Spy, Time Spy Extreme, and Port Royal have high GPU usage (they're all rendered at 1440P or higher), and most of your games have low GPU usage - yeah, CPU bound.

I believe Superposition is free benchmark, and has the ability to change resolutions. I don’t believe 1080P maxes out GPU usage, but it should be much higher than 30 to 40%.

A game that DOES scale fairly well even at 1080P is Doom Eternal. If that game exhibits high GPU usage, then you have your answer again.

When you say CPU bound, you mean limited by the cpu? If that's what you mean, I'm not sure how that can be if in game the cpu is not being pushed very hard. I'll test more games tonight and post some results. Also, I was using Heaven as the benchmark tool, is that not a good one to use?
 

just head out and buy a new MB, and one of the newer cpus that came out this year and redo the test. Then, you'll see the performance you were expecting.. 

I'm running an 8700K as well and I got this result @ 1440p

TMoseley2

DukeCLR
I'm running an 8700K as well and I got this result @ 1440p
 

Can you rerun with the following settings and post your results? These are the settings I used.
 
Preset: custom
API: DirectX11
Quality: Ultra
Tessellation: Extreme
Stereo 3D: Disabled
Multi-monitor: Disabled
Anti-aliasing: x8
Full Screen: unchecked
Resolution: System

You did see that he's running a 1440P monitor and a 3080, and that it won't be comparable to your 3070 at 1080P?

Aye, agreed on the CPU bottleneck.

Keep in mind you're about 2 and a half generations behind CPU-wise paired with a 1-month old 3070. Then you're playing on 1080p which doesn't challenge your GPU nearly as much as your CPU, because rendering at 1080p is a breeze for your 3070 while it's looking back and waiting for your CPU to catch up.

jankerson
 
 
Not just that, the 3070 is close to a 2080Ti in performance and THAT would bottleneck every CPU made at 1080P.
 

I could have sworn I mentioned that in mine as well, must have erased to retype it and.. forgot all about it  

Either way, spot on 

TMoseley2
If it's a CPU bottleneck, wouldn't the CPU be pegged out at 100% or close to it? What I'm seeing is not the case.
 
I'll rerun the benchmark this evening with MSI Afterburner running so I can see what the CPU is doing.

No. Because it depends on what tool you are using to report CPU usage. Most games don't make use of more than one core.
 
I benched Heaven using your settings at 1080p. Here is a screenshot showing nearly 100% utilization of CPU0 but next to nothing on the other cores. Overall CPU usage was 16%, but the one core running Heaven was pegged.
 
 

Trojone
Hi, I’ve just got my 3070 ftw 3 ultra this morning, and I'm getting the same kind of poor results in games as OP is in benchmarks, and I know you will say I'm CPU bound, and I won’t say that I'm not, but with the same settings in Cold War as I had on my old card, a 970,  it seems I have gained 20 -30fps. Sometimes lower. Running a benchmark in MSI afterburner, while playing cold war, I can see that the 1% low is 35 fps and the average is around 120. I have all the settings on low. I would have expected a bit more of a jump from the 970. My friend has the same CPU, i5 8600k, I have mine overclocked to 4.5, although in cold war is was only showing as 4.1 in MSI,  we have the same monitor too, 1080 144hz, both have 16gb ram, and his 2080 with the same settings was getting 50 or more fps than me.
 
This website, claims that I should be getting roughly double than I am, on ultra settings and I'm on low.
I still have a lot of games to try but so far Shadow of the Tomb Raider was half of that on the website linked above, so was BF5 and Cold War.
I have used DDU in safe mode to remove the drivers and reinstalled them a couple times.
Does anyone have any words of wisdom for me?

Try turning the game settings Up .... 1080 & low settings is not much of a load on the GPU --> that can increase the CPU bottleneck
 
Tell use more about your PC might help ... OS details, Drive for OS & Game, drivers used, PSU, RAM, Case & case fans, Room temp & GPU temp
 
Which driver are you using ? 
What Nvidia Control panel & Windows Power setting are you using ?
 
 
Might want to try a Power flush
 
      1 - Shut down
 
      2 - Unplug PC & anything connected to PC that is plugged into the wall (monitor, printer, USB hub ...)
 
      3 - Now hold "start button" in for 20 sec to discharge capacitors ....
 
      4 - Plug in & reboot
 
 
 
Long shot ... In Windows ... Turn OFF ..... the update feature "Delivery Optimization"
 
It is known to cause stuttering in Windows 10
 
 
 

Ok that’s great, thank you for your speedy reply, I will have a look at that now and come back. In the mean time, some more details on my PC.
Windows 10 version 2004.
OS Drive is m.2 SSD NVME 256gb, game drive for cold war 480gb SSD.
PSU is a be quiet straight power 750w, 16gb corsair vengeance @ 2400MHz.
Case is a be quiet dark base pro 900, 2 intake silent wings 120mm 1 exhaust silent wings 120mm.
CPU cooler is a dark rock pro 4,
Room temp is around 25 degrees, CPU temp in game is around 62 degrees , gpu temp in game is around 52 degrees.
Windows power setting is performance, as I checked this I noticed that nvidia power setting is optimal power, so I'll change that when I do the other testing you mentioned.
nvidia drivers 457.30
 
I will try all the other suggestions now and be back shortly