Low GPU usage with Crossfire R9 270x when gaming.

Toggle sidebar Toggle sidebar
P

PivMan

Reputable
May 9, 2014
12
0
4,520
I need some help here. I just installed a second R9 270x 2GB into my rig and hooked a Crossfire bridge between the two cards. Everything seems to be all fine and dandy except no matter what game I play the performance is either no better or introduces constant spikes down to thirty frames (which feels like twenty in Xfire) every ten seconds. The GPU usage for both cards are very low, usually hovering around 25%. I've tried nearly every trick in the book now and I'm about to go behead a child at this point. A few notes:

- I've disabled ULPS.
- I'm running at 1360x768 (don't ask, poor circumstances) but have also tried things at an upscaled 1080p with no difference.
- I've tried every combination of drivers installs, card reseats, and master/slave switching several times.
- I am indeed running the latest 14.4 drivers.
- I have tried disabling Vsync and it seemed to help only slightly, mostly with just stuttering.
- FurMark is giving me exactly the scores I would expect from this rig. 5000+ at 1080p.
- My mobo is running at 16x/4x, but it seems to not affect FurMark by much. I'd be content with that kind of performance.
- I've tried every clock between 3.5Ghz and 4.9Ghz.

My specs:
CPU: AMD FX-6300 w/ CM Hyper 212 EVO
GPU: Gigabyte R9 270x Crossfire
RAM: Corsair Vengeance 2x4GB 1866Mhz
MOM: ASUS M5A97 LE R2.0
PSU: EVGA SuperNOVA 750W
 
Solution
P
The usage never breaks past 93% or so per core. When one core reads ~90% another one will read 75%.

My FPS with two cards is the issue. On ultra with resolution scaling at 200% at 768p Crossfire will run 60 FPS when looking off the map/the sky while spiking down to the low fifties/high forties randomly and at 30-50 FPS when looking at the main part of the map. On the more demanding ones it tends to stay towards thirty, bump to a higher number every ten or so seconds, then back down.

On a single card I see near equal overall performance, but Crossfire has much more fluctuation and therefore may hit higher numbers more often for split seconds but it doesn't make an overall difference in playability. I should also point out that in...
Sort by date Sort by votes
P

PivMan

Reputable
May 9, 2014
12
0
4,520
Even at 768p Battlefield 4 on ultra at 200% resolution scale makes thirty FPS on a single card. My goal here is sixty. I've also tried 1080p with no change and plan on upgrading in the near future once I can afford a new monitor.

One 270x is equal to one GTX 760. Not a very high end card, but apparently kills in Crossfire.

I know full well some games don't like Crossfire but the games I test it on have plenty of support for it to make performance equal to a single card abnormal. I've been running a single card for months now and just upgraded.

My CPU usage in BF4 on stock clocks usually hits around 75%-90% across all cores, with them not each reading the same usage. Depends on whats going on around me. CPU usage in games like Skyrim are lower.
 
Upvote 0 Downvote
I

i7Baby

Titan
BANNED
Jul 25, 2012
44,084
2
106,960
CPU Usage? At 90% plus it is probably holding back the gpus.

What fps you getting with two cards in bf4 on what settings and resolution? And with one card?

Have you raised this issue on bf4 etc forums? As I said, sometimes its game settings or game client.

 
Upvote 0 Downvote
P

PivMan

Reputable
May 9, 2014
12
0
4,520
The usage never breaks past 93% or so per core. When one core reads ~90% another one will read 75%.

My FPS with two cards is the issue. On ultra with resolution scaling at 200% at 768p Crossfire will run 60 FPS when looking off the map/the sky while spiking down to the low fifties/high forties randomly and at 30-50 FPS when looking at the main part of the map. On the more demanding ones it tends to stay towards thirty, bump to a higher number every ten or so seconds, then back down.

On a single card I see near equal overall performance, but Crossfire has much more fluctuation and therefore may hit higher numbers more often for split seconds but it doesn't make an overall difference in playability. I should also point out that in Crossfire lowering graphics settings makes little to no difference in frame rate.

I've looked into Battlefield's Crossfire issues heavily and at this point in the game's release I should very well be getting some kind of improvement. Much better than this slop. Bear in mind I really only use BF4 to benchmark this stuff. I bought the second card to mod Skyrim with.

In other games I see the same scenario and with much less CPU usage. FurMark also gave me 780 tier scores so I hope it couldn't be too much of a bottleneck. I was aware the 6300 was a lowly but capable processor.

 
Upvote 0 Downvote
Solution
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom