• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Why are some 3090's only hitting ~14k GPU in Timespy bench on 3960x?

ketma

ketma

ketma

Associate
Joined
14 Dec 2003
Posts
501
Location
Lincs
As title, why are a number of 3090's only hitting 14k GPU when used on Threadripper CPU's ? i.e. 3960 and others are hitting 17-19k.

Can't be the CPU as I doubt 2-300MHz would make such a huge difference?

FPS seems to be topped out around ~90.....what's causing this discrepancy?
 

Kaapstad

Kaapstad

Kaapstad

Man of Honour
Joined
21 May 2012
Posts
31,940
Location
Dalek flagship
ketma said:
As title, why are a number of 3090's only hitting 14k GPU when used on Threadripper CPU's ? i.e. 3960 and others are hitting 17-19k.

Can't be the CPU as I doubt 2-300MHz would make such a huge difference?

FPS seems to be topped out around ~90.....what's causing this discrepancy?
Click to expand...

You need to post the results link

And yes it could be the CPU.
 

Stephen Garner

Stephen Garner

Stephen Garner

Associate
Joined
22 Sep 2020
Posts
89
Location
London, UK
A pal has a threadripper and a 3090 (he renders, and builds models and levels in Maya and Unreal), he suggests his TR doesn't get utilised fully, causing his discrepancy.

His operational use in Blender, Maya, Unreal etc are unaffected and it performs as it should.
Side note from me - my 3080 did underperform in 3DMark, but Nvidia control panel was set to Optimal Power for 3DMark, which i overloooked. Changing this sky rocketed the results.
 

OC2000

OC2000

OC2000

Associate
Joined
2 Oct 2009
Posts
789
My guess it will be down to memory too.

On the Timespy standard benchmark thread. My 3950X scores CPU 16000 with extremely tight Ram settings (3800mhz, 1900 IF CAS 16 with that Ryzen Apps tightest secondary timings)

Those with 3600 mhz are scoring around CPU 14000

And there's one that scores CPU 10000 with Ram timings of 2133
 

ketma

ketma

ketma

Associate
OP
Joined
14 Dec 2003
Posts
501
Location
Lincs
I'll post a score later as I'm using it now with work.....however from previous tests I can see the GPU load is not maxed out at all.

I think this is a Threadripper thing.
 

ketma

ketma

ketma

Associate
OP
Joined
14 Dec 2003
Posts
501
Location
Lincs
So what is the conclusion to this? Is this an issue with the benchmark that doesn't scale or use cores correctly or something else ?

Thanks.
 

Shteve

Shteve

Shteve

Associate
Joined
19 Sep 2020
Posts
72
Clearly the 24GB of memory the 3090 has isn't enough. I've no idea how the 3080 with only 10GB is going to do, and the rumoured/ cancelled 20GB 3080 will obviously also still struggle. I'm going to wait for the 58GB cards.


/s

:D
 

Kaapstad

Kaapstad

Kaapstad

Man of Honour
Joined
21 May 2012
Posts
31,940
Location
Dalek flagship
ketma said:
So what is the conclusion to this? Is this an issue with the benchmark that doesn't scale or use cores correctly or something else ?

Thanks.
Click to expand...

There is a problem with the Standard Timespy bench not using all the cores available.

With my 18/36 10980XE intel CPU I turn off Hyperthreading which gives a much better result on just 18 cores.

I have tried doing the same with my 3950X Ryzen CPU and the results are not as good, I think AMD CPUs work a bit differently.

If you really want to give your whole system a good test try running Timespy Extreme as this caters for CPUs like yours and gives a much bigger score.
 

zx128k

zx128k

zx128k

Associate
Joined
9 May 2007
Posts
1,284
With the 10980XE you need to limit the cores to 10. Take the frequency as high as you can. https://www.3dmark.com/pr/419756

Power supply calculator shows 850 watts and the build has a 1000 watt psu. Overclocking could max out the build.

Build
TR 3960X, TRX40 Aorus Pro WiFi, 32GB Corsair LPX, MSI Gaming X Trio 3090,
SX8200 Pro M.2, 2x SP P34A80 M.2, 2x MX500, 2x P300 HDD,
Corsair H115i Platinum, Corsair K70 RGB, Samsung UR59C 32" 4k,
Sound Blaster X Katana, Thermaltake View 71, HX1000i PSU.

Case can have 5x 140mm fans and 2x 120mm fans
 
Last edited:

Kaapstad

Kaapstad

Kaapstad

Man of Honour
Joined
21 May 2012
Posts
31,940
Location
Dalek flagship
zx128k said:
With the 10980XE you need to limit the cores to 10. Take the frequency as high as you can. https://www.3dmark.com/pr/419756

Power supply calculator shows 850 watts and the build has a 1000 watt psu. Overclocking could max out the build.
Click to expand...

That is a Port Royal result you have there not a Timespy Standard one.

The number of core does not matter on Port Royal, what does matter is the choice of CPU when running SLI.

The 10980XE CPUs are very poor for running SLI due to MESH they use on the cores, JayzTwoCents would have got a better score by about 500 points using a 10900k or even one of the older mainstream gaming CPUs running in a board that has SLI support.

Still that is JayzTwoCents for you.
 

ketma

ketma

ketma

Associate
OP
Joined
14 Dec 2003
Posts
501
Location
Lincs
zx128k said:
With the 10980XE you need to limit the cores to 10. Take the frequency as high as you can. https://www.3dmark.com/pr/419756

Power supply calculator shows 850 watts and the build has a 1000 watt psu. Overclocking could max out the build.

Build
TR 3960X, TRX40 Aorus Pro WiFi, 32GB Corsair LPX, MSI Gaming X Trio 3090,
SX8200 Pro M.2, 2x SP P34A80 M.2, 2x MX500, 2x P300 HDD,
Corsair H115i Platinum, Corsair K70 RGB, Samsung UR59C 32" 4k,
Sound Blaster X Katana, Thermaltake View 71, HX1000i PSU.

Case can have 5x 140mm fans and 2x 120mm fans
Click to expand...

I'm not overclocking to keep within a safe range of the PSU.
 
You must log in or register to reply here.
Back
Top Bottom