R9 380x under-performing (poor GPU usage and FPS drops)

njbrugh

Commendable
Feb 20, 2016
3
0
1,510
Hello all,

So I bought a new R9 380x within the last week. As soon as I installed it I download Rise of the Tomb Raider, because you know... graphics. Anyway, things went well for the first 10 minutes of the game, I was getting around 50 - 60 fps on high preset with shadows turned down to medium, but then when I got to the Siberian Wilderness things got weird. I was getting 40 - 50 fps which I was fine with however I would get fps drops to below 30 which made it very hard to play. So I pulled up msi afterburner and noticed that my gpu usage was all over the place (see pics below). So I decided to experiment to see if I could fix things and here's what I tried so far.

- Turning vsync off
- Turning graphics preset down and even at the lowest setting I was getting around the same fps (maybe 5 - 10 fps better) and the usual stuttering and fps drops
- Clean installing drivers, multiple times, (versions 16.1.1 and 15.12)
- Different games Firewatch (was getting the same stutter but overall ran better); and The Division beta high preset (almost worse than Rise of the Tomb Raider averaging 35 - 40 fps)
- Reinstalling old graphics card R9 270 (oddly enough, in Rise of the Tomb Raider I was able to achieve the same fps at medium settings as I did with the 380x and the game play was much more smooth, no horrible fps drops and stuttering)
- Turned windows power mode from balanced to high performance
- Running furmark (which on the 1080p preset I was able to achieve 100% gpu usage the whole time with the average fps of 52)
- Purchased new psu; Corsair CX850M
- Overclocking CPU (4.4 Ghz)
- Checked GPU temps they don't get above 70c

Specs:
CPU - AMD FX-6300 (oc to 4.4 Ghz)
CPU Cooling - Corsair H60 water
RAM - 24Gb 1600mhz
MB - Gigabyte GA-970A-D3P
PSU - Corsair CX850M

Below is msi afterburner pic while playing Rise of the Tomb Raider

xUb25rj.png


I started the game about half way through as you can see by the sharp movement of the GPU usage. The steady 90% - 100% usage that you see towards the end was when I went into a small cave and ran around.

Sorry for the wall of text but I wanted to make sure I told the whole story. At this point I'm pretty stumped. Does anybody know what could be going on or do you have any solutions I could try?

I'm leaning towards either a bad GPU or CPU bottle necking which I somewhat doubt especially with 4.4ghz oc. I would like to avoid returning the GPU if at all possible because I purchased it from Amazon and I would rather not pay their restocking fee.

Any help would be appreciated. Thanks!
 
It does sounded like faulty card so far. But what about other games? Does it also suffer performance loss? If the problem is app specific than something else might be at fault here.

What's odd to me is; if it can sustain furmark bench without hiccups or artifact it should be fine.

Another possibility is the driver AMD provide hasn't optimize enough for that game. But that doesn't explain lots of people posting on youtube with similar card running RoTB just fine.

My suggestion would be, ever try it on another PC? Just to see if that odd behavior persist.
 


Thanks for the response. Yeah I've tried some other games, in the process of trying more, unfortunately it seems to be happening on all of them. Just tried Heaven Benchmark and it did fine. I had constant fps the whole time. My buddy has a tower so maybe he'll let me try it in his.
 
Ok so I just played a good bit of Bioshock Infinte on Ultra settings. With vsync of I was getting around 120 - 130 fps while just walking around and gpu usage was normal. However when I would get into action the fps would drop down to around 90 sometimes even 80 and the gpu usage would start jumping around again. Because the fps are so high in the first play it doesn't effect the gameplay however I feel like the fps drop is still causing some micro stutter. Could all of this be signs pointing to CPU problems, and if so how can I go about testing it?
 
Hey njbrugh!

The problem is caused by Crimson drivers. It has some serious downclocking issues. Try latest catalyst drivers (15.11.1 Beta) and let us know. I have the same card and I had the same problem. Until they fix downclocking issues I would stick with catalyst drivers.

Cheers!
 
So, I know that this question is old, but I thought I'd try to help either way. This is just a monitoring issue that you have here with MSI Afterburner. I'd had the same issue myself, and I had spent months pulling out my hair over it until I finally came across an answer that worked.

Open up Settings in Afterburner, and under the General tab, and all the way at the bottom, you'll see "enable unified GPU usage monitoring". Check that, and you should see your GPU usage stabilize in Afterburner. Hopefully that helps you out!

Hell, I'm not even sure you have the card anymore, but maybe someone looking for the answer to this will run across this as well. 😀