60% usage in Far Cry 4 GTX 970 SLI

Restitution

Reputable
May 8, 2014
47
0
4,530
My SLI set-up won't go above 60% usage, 70% max. This is causing fps issues in some areas like towns, and random fps drops. Anyone know why my cards won't put out all they can muster? They just lazy?
 
Full system specs...

Motherboard Make Model = ?
GFX Card Make / Model = ?
PSU Make / Model = ?
Storage Make / Model = ?
RAM Make / Model / Speed / CAS = ?

Is CPU Overclocked ?
Are GPUs Overclocked ? and if so what settings ?


Have you tested with Furmark ?
-What is reported TDP ?
-What is reported Temp after 20-30 minutes ?
-Core Clock = ?
-Memory Clock = ?
-Max Boost Clock = ?
 
If you wanted detailed system specs then please be specific, I appreciate the help though.

Motherboard: ASUS Z97
GPU: EVGA SSC GTX 970
PSU: NZXT Hale 90 v2
Storage: 500GB Samsung 840 Evo SSD
RAM: Corsair Vengeance Low Profile 2x4 GB

No overclocked settings.

I can get back to you on a furmark test.
 


"System specs" include all items in the system. It's impossible to judge potential problems with GPU for example not using enough power without the basic information such as how much power is available, how well the unit delivers it and whether the system is overclocked. To judge the ability of the PSU, we need to know the wattage of the PSU. We still don't know that as the NZXT Hale 90 v2 comes in various wattages....but since the smallest one I am aware of is 850 watts we can rule out PSU quality and wattage as an issue (assuming it's not a bum unit).

GFX card performance is impacted by RAM Speed and CAS to a minor extent (tho up to 10% in some games like F1) but this becomes more of a factor in SLI....so RAM speed and CAS is useful information to have and why I asked for it, but I doubt it's our problem here with 8GB.

Searching newegg on Asus Z97 brings up 36 products...Z97-A, Z97-Pro, Z97E, Z97-I, Z97M and so on.... Which one you are using is important because boards with a 3rd PCI=E slot will drop from x8 to x4 if you stick a card in that 3rd slot such as the Asus RAIDR Express.

As for the card, we have conflicting information in that Furmark is showing a lot more usage (up to 88% / 98%) than you're seeing under FC4. That tells me that for whatever reason FC4 isn't asking for that much power. 1400 max boost clock is a bit low....temps are high at 80C as these are normally in mid 60s for the 970. You said this was the SSC not the SC so if that's the case, we don't have to worry about the 3rd heat pipe missing the HS. the original SC also ....

On the other side of the GPU is a metal contact plate that partially cools two of the four memory chips on this side, leaving the other two exposed. It also cools the MOSFETs of the power phases serving the memory, but no thermal pads are used, so heat transfer is likely to be limited.

I don't know if this design is carried thru on the SSC. So the thing that stands out is "Is the card getting enough cooling .... not necessarily to the GPU but the VRMs, MOFSETS and memory. Knowing the case and number / rpm of fans might give a clue .... also, if you have the means, case interior temp and ambient might give a clue.

I'm wasn't really worried about resolution as the question for me is not fps but why it's not using available power. However we see that you are getting near 100% power usage in Furmark... the 2nd card is usually be a bit behind. My 1st card usually runs about 112%... the 2nd about 108%. Why FC4 is not asking for more is therefore puzzling.

In an attempt to narrow the cause down, Id adjust the fan curve to max and see if it makes a difference. Also swap card positions and see if one shows any deficiency compared to the other......
 

Hmm, nice. Except for the game does a pretty bad job at it, but I suppose that's Ubisoft's fault, and I rarely laugh when I'm here :)
 


There were issues with Far Cry 3, since resolved, but Far Cry 4 runs fine on my box. My son plays FC4 on my box (twin WC'd 780s) and he's not seeing any issues and the reviewers aren't seeing it. My son said he sees fps in low 60s w/ one card and 110+ in SL.

But lest see what the test sites say .....

http://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_960_SLI/14.html

Scaling in Far Cry 4 @ 1920 x 1080 = 84%
Scaling in Far Cry 4 @ 2560 x 1440 = 94%

That's looks pretty darn good to me considering the average scaling over TPUs 20 game test suite:

http://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_960_SLI/23.html

Average Scaling in 20 games @ 1920 x 1080 = 52%
Average Scaling in 20 games @ 1920 x 1080 = 54%

That works out to FC4's scaling being:

62% better than average at @ 1920 x 1080
74% better than average at @ 2560 x 1440

I just don't understand how that can be classified as a "pretty bad job".... I consider 84-94% scaling to be excellent.
 


Fascinating, you earned a medal.
 
Did you swap card positions.... change fan curve ?

Install MSI Afterburner and look at your power limit %

index.php