Jump to content

R9 290 low gpu usage

THGM

Would you like to run quick cinebench openGL test and show us graph(gpu usage) from msi afterburner? :)

And yea, i feel pretty scammed with my R9 290 considering this issue, much worse than the issue with the 970 for example. Shame on AMD and reviewers they missed out that important part.

Its my bad man, i didn't mentioned its about 3, but now that i think of it, i'll contact ea to give me some gimetime for bf4 and see what happens there.

cinebench isnt a good gpu benchmark its simply not demanding enough to use gpu's like this, before saying this sint true would you expect it to use quadfire 290's?

Falcon: Corsair 750D 8320at4.6ghz 1.3v | 4GB MSI Gaming R9-290 @1000/1250 | 2x8GB 2400mhz Kingston HyperX Beast | Asus ROG Crosshair V Formula | Antec H620 | Corsair RM750w | Crucial M500 240GB, Toshiba 2TB, DarkThemeMasterRace, my G3258 has an upgrade path, my fx8320 doesn't need one...total cost £840=cpu£105, board£65, ram£105, Cooler £20, GPU£200, PSU£88, SSD£75, HDD£57, case£125.

 CASE:-NZXT S340 Black, CPU:-FX8120 @4.2Ghz, COOLER:-CM Hyper 212 EVO, BOARD:-MSI 970 Gaming, RAM:-2x4gb 2400mhz Corsair Vengeance Pro, GPU: SLI EVGA GTX480's @700/1000, PSU:-Corsair CX600m, HDD:-WD green 160GB+2TB toshiba
CASE:-(probably) Cooltek U1, CPU:-G3258 @4.5ghx, COOLER:-stock(soon "MSI Dragon" AiO likely), BOARD:-MSI z87i ITX Gaming, RAM:-1x4gb 1333mhz Patriot, GPU: Asus DCU2 r9-270 OC@1000/1500mem, PSU:-Sweex 350w.., HDD:-WD Caviar Blue 640GB
CASE:-TBD, CPU:-Core2Quad QX9650 @4Ghz, COOLER:-OCZ 92mm tower thing, BOARD:-MSI p43-c51, RAM:-4x1GB 800mhz Corsair XMS2, GPU: Zotac GTX460se @800/1000, PSU:-OCZ600sxs, HDD:-WD green 160GBBlueJean-A
 CASE:-Black/Blue Sharkoon T9, CPU:-Phenom2 x4 B55 @3.6Ghz/1.4v, COOLER:-FX8320 Stock HSF, BOARD:-M5A78L-M/USB3, RAM:-4GB 1333mhz Kingston low profile at 1600mhz, GPU:-EVGA GTX285, PSU:-Antec TP550w modu, STORAGE:-240gb  M500+2TB Toshiba
CASE:-icute zl02-3g-bb, CPU:-Phenom2 X6 1055t @3.5Ghz, COOLER:-Stock, BOARD:-Asrock m3a UCC, RAM:2x2GB 1333mhz Zeppelin (thats yellow!), GPU: XFX 1GB HD6870xxx, PSU:-some 450 POS, HDD:-WD Scorpio blue 120GB
CASE:-Packard Bell iMedia X2424, Custom black/red Aerocool Xpredator fulltower, CPU's:-E5200, C2D [email protected]<script cf-hash='f9e31' type="text/javascript"> /* */</script>(so e8500), COOLER:-Scythe Big shuriken2 Rev B, BFG gtx260 sp216 OC, RAM:-tons..
Gigabyte GTX460, Gigabyte gt430,
GPU's:-GT210 1GB,  asus hd6670 1GB gddr5, XFX XXX 9600gt 512mb Alpha dog edition, few q6600's
PICTURES CASE:-CIT mars black+red, CPU:-Athlon K6 650mhz slot A, COOLER:-Stock, BOARD:-QDI Kinetiz 7a, RAM:-256+256+256MB 133mhz SDram, GPU:-inno3d geforce4 mx440 64mb, PSU:-E-Zcool 450w, STORAGE:-2x WD 40gb "black" drives,
CASE:-silver/red raidmax cobra, CPU:-Athlon64 4000+, COOLER:-BIG stock one, BOARD:-MSI something*, RAM:-(matched pair)2x1GB 400mhz ECC transcend, GPU:-ati 9800se@375core/325mem, PSU:-pfft, HDD:-2x maxtor 80gb,
PICTURES CASE:-silver/red raidmax cobra (another), CPU:-Pentium4 2.8ghz prescott, COOLER:-Artic Coolering Freezer4, BOARD:-DFI lanparty infinity 865 R2, RAM:-(matched pair)2x1GB 400mhz kingston, GPU:-ati 9550@375core/325mem, PSU:-pfft, HDD:-another 2x WD 80gb,
CASE:-ML110 G4, CPU:-xeon 4030, COOLER:-stock leaf blower, BOARD:-stock raid 771 board, RAM:-2x2GB 666mhz kingston ECC ddr2, GPU:-9400GT 1GB, PSU:-stock delta, RAID:-JMicron JMB363 card+onboard raid controller, HDD:-320gb hitachi OS, 2xMaxtor 160gb raid1, 500gb samsungSP, 160gb WD, LAPTOP:-Dell n5030, CPU:-replaced s*** cel900 with awesome C2D E8100, RAM:-2x2GB 1333mhz ddr3, HDD:-320gb, PHONE's:-LG optimus 3D (p920) on 2.3.5@300-600mhz de-clock (batteryFTW)
Link to comment
Share on other sites

Link to post
Share on other sites

Yea i though its only for amd that way in cinebench.

For me its about 50% gpu usage on r9 290 and 126 fps, up to 130fps sometimes so its not "locked", but you can try some game benchmarks such as hitman absolution, batman etc. so we can compare better

Well i am playing dragon age inquisition at the moment (high settings, HBAO full, 2x MSAA) , i am running at 98% gpu usage at the moment (50-55fps)

I refuse to read threads whose author does not know how to remove the caps lock! 

— Grumpy old man

Link to comment
Share on other sites

Link to post
Share on other sites

You dont need to quote everything, try snipping.

 

what resolution?

this gpu is meant for 1440p+ im not surprised if its at 900p..

he is running at 1680x1050...that card is way overkill for that resolution.

I refuse to read threads whose author does not know how to remove the caps lock! 

— Grumpy old man

Link to comment
Share on other sites

Link to post
Share on other sites

he is running at 1680x1050...that card is way overkill for that resolution.

Yeah that's it, maybe if he runs super sampling then it'll be used.

if you want really high fps with a really high end card with a not very high resolution you need high ipc and high clock speed (and depending on engine, lots of threads)

so a 4690k overclocked will kill anything in bf4 at 720p with a 290 and mantle and get say "80fps" (just for example)

but if you were to stream that gameplay the xeon here with its slower 8 threads will maybe be at 90% of that speed so 70fps but the i5 will struggle and maybe get 40.. thats where the differences are.

hitman absolution has been proven to be a terrible game for both cpu usage AND for nvidia (will find it later but you can find it yourself)

and metro is a badly coded bitch too, great looking though.

crysis3 is a well threaded game and 90% towards being perfect as it is gpu bound, but the cpu scheduling could still do with some polishing, but this wont happen and isnt something a driver can fix.

I never had problems with amd drivers but I am liking the new way amd are looking towards OMEGA and the care they're putting into them, frame pacing is now something of the past and dynamic super resolution is something op could use to make use of his card.

anyone want to me to do some tests to verify things? i have bf4, crysis3, COD AW (i think somewhere).

you can see the coralation between cpu speed and cores here, this isnt great

http://www.techspot.com/review/608-hitman-absolution-performance-benchmarks/page6.html

I put money on it using every other thread (so 3 with an fx6) and not using hyperthreading worth a damn either.

oh and alien isolation isn't demanding, my friends playing it on 1680x1050 averaging 60fps on high.....with an athlon 640 and gtx460.

http://www.techspot.com/review/903-alien-isolation-benchmarks/page2.html

http://www.techspot.com/review/903-alien-isolation-benchmarks/page5.html

Falcon: Corsair 750D 8320at4.6ghz 1.3v | 4GB MSI Gaming R9-290 @1000/1250 | 2x8GB 2400mhz Kingston HyperX Beast | Asus ROG Crosshair V Formula | Antec H620 | Corsair RM750w | Crucial M500 240GB, Toshiba 2TB, DarkThemeMasterRace, my G3258 has an upgrade path, my fx8320 doesn't need one...total cost £840=cpu£105, board£65, ram£105, Cooler £20, GPU£200, PSU£88, SSD£75, HDD£57, case£125.

 CASE:-NZXT S340 Black, CPU:-FX8120 @4.2Ghz, COOLER:-CM Hyper 212 EVO, BOARD:-MSI 970 Gaming, RAM:-2x4gb 2400mhz Corsair Vengeance Pro, GPU: SLI EVGA GTX480's @700/1000, PSU:-Corsair CX600m, HDD:-WD green 160GB+2TB toshiba
CASE:-(probably) Cooltek U1, CPU:-G3258 @4.5ghx, COOLER:-stock(soon "MSI Dragon" AiO likely), BOARD:-MSI z87i ITX Gaming, RAM:-1x4gb 1333mhz Patriot, GPU: Asus DCU2 r9-270 OC@1000/1500mem, PSU:-Sweex 350w.., HDD:-WD Caviar Blue 640GB
CASE:-TBD, CPU:-Core2Quad QX9650 @4Ghz, COOLER:-OCZ 92mm tower thing, BOARD:-MSI p43-c51, RAM:-4x1GB 800mhz Corsair XMS2, GPU: Zotac GTX460se @800/1000, PSU:-OCZ600sxs, HDD:-WD green 160GBBlueJean-A
 CASE:-Black/Blue Sharkoon T9, CPU:-Phenom2 x4 B55 @3.6Ghz/1.4v, COOLER:-FX8320 Stock HSF, BOARD:-M5A78L-M/USB3, RAM:-4GB 1333mhz Kingston low profile at 1600mhz, GPU:-EVGA GTX285, PSU:-Antec TP550w modu, STORAGE:-240gb  M500+2TB Toshiba
CASE:-icute zl02-3g-bb, CPU:-Phenom2 X6 1055t @3.5Ghz, COOLER:-Stock, BOARD:-Asrock m3a UCC, RAM:2x2GB 1333mhz Zeppelin (thats yellow!), GPU: XFX 1GB HD6870xxx, PSU:-some 450 POS, HDD:-WD Scorpio blue 120GB
CASE:-Packard Bell iMedia X2424, Custom black/red Aerocool Xpredator fulltower, CPU's:-E5200, C2D [email protected]<script cf-hash='f9e31' type="text/javascript"> /* */</script>(so e8500), COOLER:-Scythe Big shuriken2 Rev B, BFG gtx260 sp216 OC, RAM:-tons..
Gigabyte GTX460, Gigabyte gt430,
GPU's:-GT210 1GB,  asus hd6670 1GB gddr5, XFX XXX 9600gt 512mb Alpha dog edition, few q6600's
PICTURES CASE:-CIT mars black+red, CPU:-Athlon K6 650mhz slot A, COOLER:-Stock, BOARD:-QDI Kinetiz 7a, RAM:-256+256+256MB 133mhz SDram, GPU:-inno3d geforce4 mx440 64mb, PSU:-E-Zcool 450w, STORAGE:-2x WD 40gb "black" drives,
CASE:-silver/red raidmax cobra, CPU:-Athlon64 4000+, COOLER:-BIG stock one, BOARD:-MSI something*, RAM:-(matched pair)2x1GB 400mhz ECC transcend, GPU:-ati 9800se@375core/325mem, PSU:-pfft, HDD:-2x maxtor 80gb,
PICTURES CASE:-silver/red raidmax cobra (another), CPU:-Pentium4 2.8ghz prescott, COOLER:-Artic Coolering Freezer4, BOARD:-DFI lanparty infinity 865 R2, RAM:-(matched pair)2x1GB 400mhz kingston, GPU:-ati 9550@375core/325mem, PSU:-pfft, HDD:-another 2x WD 80gb,
CASE:-ML110 G4, CPU:-xeon 4030, COOLER:-stock leaf blower, BOARD:-stock raid 771 board, RAM:-2x2GB 666mhz kingston ECC ddr2, GPU:-9400GT 1GB, PSU:-stock delta, RAID:-JMicron JMB363 card+onboard raid controller, HDD:-320gb hitachi OS, 2xMaxtor 160gb raid1, 500gb samsungSP, 160gb WD, LAPTOP:-Dell n5030, CPU:-replaced s*** cel900 with awesome C2D E8100, RAM:-2x2GB 1333mhz ddr3, HDD:-320gb, PHONE's:-LG optimus 3D (p920) on 2.3.5@300-600mhz de-clock (batteryFTW)
Link to comment
Share on other sites

Link to post
Share on other sites

I knew that this card is overkill for my resolution, i just wanted futureproof gpu which will last at least few years, but in this case if theres still issues with most games it wont last more than few months for me.

About the gpu usage - in most games, when the gpu usage isn't at full 99% i seem to be getting the same FPS on different resolutions - 640x480(40-50% gpu usage) vs 1680x1050(60-70%), same goes if i put super sampling, then gpu usage might hit 99% or keep at it, but sometimes fps will be either lower(usual) or the same as before, which is in most, if not all cases.

Heres the result after playing bf4 multiplayer on ultra with 2xMSAA:
LfNWArB.png

Link to comment
Share on other sites

Link to post
Share on other sites

Firstly, bf4 multiplayer is more cpu intensive than gpu. Do you get the same thing during single player?

 

And as you pointed out, the lower the resolution the less your card is being used and the more its becoming cpu dependant. As you scale upward to the 1680x1050 and turn on more sampling your card is being used to its potential. The card seems to be doing fin with a minimum fps of 80ish. Not really seeing a huge issue with what you are describing so far. At the resolution you are working worth, that card will only be used at full with high sampling/aa or higher resolution.

 

edit: lastly - what kind of monitor? 60hz or higher?

 

double edit: your fps will fluctuate depending on whats on screen. It will never be 100% at max fps all the time (just in case you werent aware)

I refuse to read threads whose author does not know how to remove the caps lock! 

— Grumpy old man

Link to comment
Share on other sites

Link to post
Share on other sites

I knew that this card is overkill for my resolution, i just wanted futureproof gpu which will last at least few years, but in this case if theres still issues with most games it wont last more than few months for me.

About the gpu usage - in most games, when the gpu usage isn't at full 99% i seem to be getting the same FPS on different resolutions - 640x480(40-50% gpu usage) vs 1680x1050(60-70%), same goes if i put super sampling, then gpu usage might hit 99% or keep at it, but sometimes fps will be either lower(usual) or the same as before, which is in most, if not all cases.

Heres the result after playing bf4 multiplayer on ultra with 2xMSAA:

LfNWArB.png

What's your fps in BF4? With my rig (details in signature) I get a solid 100 fps on most conquest large maps with ultra preset. GPU usage is always around 98% for me. I use a 144Hz 1080p display and even when using 1440p downsampled I get 80 fps.

I believe it's either your CPU that's holding you back or bad drivers/GPU.

Mah Build:  CPU: Intel i7-7700K COOLING: NZXT Kraken X52 GPU: Asus Strix GTX 1080 MOBO: Asus Strix Z270F RAM: 16GB GSKILL TridentZ RGB 3000 

CASE: In Win 303 STORAGE: Samsung 850 EVO 250GB, Seagate 1TB  SSHD PSU: EVGA Supernova G2 750W

Link to comment
Share on other sites

Link to post
Share on other sites

I'm playing BF4 on ultra preset(4x msaa and on mantle) on 1680x1050, 120hz and vsync off. On some maps im getting 100-120fps, but on some maps with much effects(rain, building destroying etc) im getting 80-90.

I'm aware FPS isnt static and depends on your location, but to repeat myself - isn't gpu usage supposed to be always at 99% so the frames are maxed out?

Gpu usage is pretty much like this in bf4 multiplayer(fps depends on the map as i said above):

NNjZaJ8.png

But i don't think this is very reliable(it all depends on map, players, gamemode etc etc), maybe we should compare with actual game benchmark.

Link to comment
Share on other sites

Link to post
Share on other sites

I'm playing BF4 on ultra preset(4x msaa and on mantle) on 1680x1050, 120hz and vsync off. On some maps im getting 100-120fps, but on some maps with much effects(rain, building destroying etc) im getting 80-90.

I'm aware FPS isnt static and depends on your location, but to repeat myself - isn't gpu usage supposed to be always at 99% so the frames are maxed out?

Gpu usage is pretty much like this in bf4 multiplayer(fps depends on the map as i said above):

NNjZaJ8.png

But i don't think this is very reliable(it all depends on map, players, gamemode etc etc), maybe we should compare with actual game benchmark.

Pretty spot on from my encounters..FPS wise I mean,.. I have the same 90-99% GPU usage in BF4 1080p Ultra DX11 in Online matches, also seen with MSI.

Maximums - Asus Z97-K /w i5 4690 Bclk @106.9Mhz * x39 = 4.17Ghz, 8GB of 2600Mhz DDR3,.. Gigabyte GTX970 G1-Gaming @ 1550Mhz

 

Link to comment
Share on other sites

Link to post
Share on other sites

I'm playing BF4 on ultra preset(4x msaa and on mantle) on 1680x1050, 120hz and vsync off. On some maps im getting 100-120fps, but on some maps with much effects(rain, building destroying etc) im getting 80-90.

I'm aware FPS isnt static and depends on your location, but to repeat myself - isn't gpu usage supposed to be always at 99% so the frames are maxed out?

Gpu usage is pretty much like this in bf4 multiplayer(fps depends on the map as i said above):

NNjZaJ8.png

But i don't think this is very reliable(it all depends on map, players, gamemode etc etc), maybe we should compare with actual game benchmark.

 

CPU is often called the brains of the PC, and GPU is essentially a laborer. The CPU sets up the frames and passes them to the GPU to render. When the GPU usage is below 99% it means it has extra processing power that's not being used. It could use it to render more frames, but doesn't usually because CPU is the bottleneck. It could also use it to render more pixels if you increase resolution.

Now, you can always create a CPU bottleneck by lowering resolution or getting more GPU horsepower, because you are shifting bottleneck to the CPU that now has to push a lot more frames. The question is, are you satisfied with the amount of fps at the point where CPU hits its limit?

 

So if you're aiming for as much fps as possible, your CPU will dictate the amount of fps. It will work full load while your GPU wouldn't. You can then either OC your CPU to try to squeeze out some more fps out of it, or switch to Nvidia to get DX11 driver with less CPU overhead. Or, increase resolution/AA to make use of the extra GPU power just for eye candy.

i7 9700K @ 5 GHz, ASUS DUAL RTX 3070 (OC), Gigabyte Z390 Gaming SLI, 2x8 HyperX Predator 3200 MHz

Link to comment
Share on other sites

Link to post
Share on other sites

Have you tried using mantle to ease up some of the CPU burden as a test?

I refuse to read threads whose author does not know how to remove the caps lock! 

— Grumpy old man

Link to comment
Share on other sites

Link to post
Share on other sites

CPU is often called the brains of the PC, and GPU is essentially a laborer. The CPU sets up the frames and passes them to the GPU to render. When the GPU usage is below 99% it means it has extra processing power that's not being used. It could use it to render more frames, but doesn't usually because CPU is the bottleneck. It could also use it to render more pixels if you increase resolution.

Now, you can always create a CPU bottleneck by lowering resolution or getting more GPU horsepower, because you are shifting bottleneck to the CPU that now has to push a lot more frames. The question is, are you satisfied with the amount of fps at the point where CPU hits its limit?

 

So if you're aiming for as much fps as possible, your CPU will dictate the amount of fps. It will work full load while your GPU wouldn't. You can then either OC your CPU to try to squeeze out some more fps out of it, or switch to Nvidia to get DX11 driver with less CPU overhead. Or, increase resolution/AA to make use of the extra GPU power just for eye candy.

Funny thing is that cpu usage is about 30-40% on each core... I'm thinking its most likely amd drivers not being as effective as nvidia ones, hence the "bottleneck". And yea im using mantle on BF4

Link to comment
Share on other sites

Link to post
Share on other sites

Could it be an issue with your card? Maybe test it out in a friend's build if possible. :)

 

Edit: I had an issue like that with my Mobo and gtx 660. I went to a friend's house and used his PC for my card and found out it was my Mobo's PCI slot. I know it isn't the motherboard but easy way to trouble shot.

01101100 01110110 01101100 00100000 00110101 00100000 01101101 01100001 01100111 01101001 01101011 01100001 01110010 01110000 

Link to comment
Share on other sites

Link to post
Share on other sites

I'm getting more confused :( Some people say its issue and some people say its usual for amd.

Link to comment
Share on other sites

Link to post
Share on other sites

I'm getting more confused :( Some people say its issue and some people say its usual for amd.

 

That's why I said see if you can test it in another build with same drivers. :D

01101100 01110110 01101100 00100000 00110101 00100000 01101101 01100001 01100111 01101001 01101011 01100001 01110010 01110000 

Link to comment
Share on other sites

Link to post
Share on other sites

I still don't see the problem the op is talking about. At 1680x1050 he will never push that card to 100% usage. It's too powerful for that resolution. He said it himself when he turns on a lot of sampling/aa it hits 100%. This is expected. His frame rates are over 80 fps which I suspect has to do with bf4 multiplayer and his CPU . Everything seems ok.

He should never hit 100% until he stresses the card to what it can do.

I refuse to read threads whose author does not know how to remove the caps lock! 

— Grumpy old man

Link to comment
Share on other sites

Link to post
Share on other sites

Funny thing is that cpu usage is about 30-40% on each core... I'm thinking its most likely amd drivers not being as effective as nvidia ones, hence the "bottleneck". And yea im using mantle on BF4

 

I'm getting more confused :( Some people say its issue and some people say its usual for amd.

 

You misunderstood. Mantle is fine. It's AMD's DX11 driver that is significantly less efficient than Nvidia's. 
You said you were getting like 80 fps on some maps while GPU usage wasn't pegged at 99%. You also said lowering the resolution lowers GPU usage. That's normal. At 1680x1050 with a 290 you're CPU bound. You wouldn't see 99% GPU usage with a GTX 970 either, although due to more efficient driver you'd see more GPU usage and more fps.  
 
GPU usage does not have to be at 99%. In fact, it's perfectly normal for it not to be at 99% at your resolution. 

i7 9700K @ 5 GHz, ASUS DUAL RTX 3070 (OC), Gigabyte Z390 Gaming SLI, 2x8 HyperX Predator 3200 MHz

Link to comment
Share on other sites

Link to post
Share on other sites

Alright, thanks a lot! The topic can be considered solved.

Link to comment
Share on other sites

Link to post
Share on other sites

  • 3 weeks later...

Just to check in again:

I tried killing floor on maximum settings(highest), 1680x1050 and i get like 20% gpu usage and about 40-80fps on average(vsync is on and forced on 120hz), i get same fps even on 1280x720. Usual?

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×