Hello gamesas!

Recently, I've acquired a Radeon 7970 Ghz edition GPU. Well, I am having a slight problem with it. For the most part, the GPU caps out my frame rate at 125 FPS (set in my autoexec.cfg). But, almost randomly, it drops to ~35 FPS for a few seconds (like 15-20) and then jumps back up. Current running the Catalyst 13.5 Beta 2 drivers. Any idea on how to fix this?

Specs:

CPU: Intel Core i5 750 @ 4.0Ghz
GPU: Radeon 7970 Ghz 1150 Mhz Core Clock 1500 Mhz Memory Clock
RAM: 8GB 1666Mhz Mushkin Blackline
Mobo: ASUS P755D Pro
HDD: 7,200 RPM Seagate
PSU: 750 TX Corsair
SSD: Adata S511 120GB

I had nasty FPS drops with V sync on with my 7950 is yours on ?

No sir, it is turned off!

I had best single card performance with 13.2 beta 6, I haven't tested single card performance since then. I like the latest drivers because they copy nvidia and make stealth a bit easier to see... ... ... I hate nvidia users in C3 that aren't affected by stealth >.>.


If the drop is in single player then its your CPU.


If you were in Australia I could give you a i7 920 and lga1336 board to test XD. If you were using crossfire that cpu would definately be bottlenecking you (battlenecking) but with a single card... hmmm.

I assume you have got your c3 on your SSD. Do you have pagefile disabled? I get randomness when theres a memory leak and it tries to write to pagefile which I have mostly disabled, so I had to buy more ram. However thats with crossfire. However at that high frame rate, what are the temps on that card, it could be getting close to 80-90 depending which card you have and case and that could be throttling the gpu. Though you have probably checked that already.

I was getting that random instability when I had google chrome in the background, if that counts for anything XD.

You may not like this. I had two of the XFX 7970's and got rid of them and will not purchase from Newegg or XFX ever again.

Anyway, here is the conversation I had with XFX support.

[ 1/27/2013 11:40:10 PM] XFX Support, I bought two of the FX-797G-TDFC and they are defective when the cards are ran at 100% GPU usage. Stock settings GPU @ 1050mhz, Memory @ 1500mhz, and Power control setting @ 0%. It does not matter whether the cards are in CrossFireX enabled mode or Disabled mode. Have also tried this with (both cards) just a single card in the system at a time. Here is what happens. The card (s) runs great until in hits 100%. At the moment of 100% the GPU core clock starts osculating from 537mhz to 1000mhz. Here is a list of programs I have used and it happens every time it hits 100% GPU usage: 1. Heaven 3.0 2. 3DMark 11 3. 3D Vantage 4. Windows Experience 5. Crysis 1 6. Crysis 2 7. Battle Field 3 8. PC Mark 7 9. FurMark 1.10.4 FurMark (Benchmark preset: 1080) is the quickest program to run to produce the same results. I first noticed this when running GPUShark v0.6.7 to monitor the card (s). Also used GPUZ v0.6.7 and logged to file to verify the results. Then loaded MSI AfterBurner to monitor and log the results and this is where the numbers from 537mhz to 1000mhz showed up. The core clock is very erratic once the GPU hit 100% and the VDDC was a constant 1.213v. At first I thought it might be a driver issue. Started with Catalyst 13.1, then moved down to Cat 12.10, then down to 12.8 Every time I uninstalled a driver, I ran the AMD Cleanup util 1.2.1.0 and Driver Sweeper and then ran CCleaner to remove any leftover registry entries. System information: Windows 7 64bit Ultimate with latest patches and updates Mother Board: Asus Rampage IV Extreme Bios v3301 Processor: Intel I7 3930K Ram: 32 gig DDR3 2134 Gskill Power Supply: SilverStone Olympus 1500 VGA Card 1: in slot 1 or BusID 1 FX-79G-TDFC Serial NXY000670 Reg Code: 0000 VGA Card 2: in Slot 3 or BusID 3 FX-79G-TDFC Serial NXY000937 Reg Code: 0000

[KEVIN_C 1/28/2013 5:03:34 PM] Ok so with just one 7970 in the computer with no other video card in the system it becomes unstable in furmark? Does it actually show any gliching or crashing or anything outside of the numbers of the clock speeds changing?

[ 1/28/2013 5:48:48 PM] Hello, thanks for the quick reply back. Yes with only one card in the system furmark runs and as soon as it does the core starts osculating. There is no glitching, crashing other than just the numbers of the clock speed. I have even set the cores at 900mhz (in crossfire) and it does the same thing. Temps max while running furmark 61 and 56c. I will try this again when I get home this afternoon and run each card individually.

[KEVIN_C 1/28/2013 6:35:58 PM] Sounds like the card is functioning as it`s supposed to then, the clock speed is set to go to 1050 at `max`, that doesn`t mean it will be running that way at all times and if you play a game it will vary depending what you are doing, how stressful a given environment is etc..

[ 1/28/2013 7:29:23 PM] During all games BF3, Crysis1 and Crysis 2, I can see the gpu clock (on second monitor) pegged at 99% and the core clock is dead on 1050mhz, but the second it hits 100% it drops. So this is the way the cards are supposed to function, in Crossfire or single?

[KEVIN_C 1/28/2013 7:47:55 PM] Hmm, yeah that does sound like it could be an issue and I`m surprised it actually even goes to 100% but all 3 of the games you mentioned are some of the most graphically intensive games ever created. Just to make sure not just one of the cards but either of the cards cause the same thing when by itself in the system right? My question is if the fluctuation in these numbers are actually causing any real problem in games or in the system or you are just curious about numbers you are seeing?

[ 1/29/2013 12:21:31 AM] Hello, I ran the test again in single card configuration. This is how I ran it. First set the core clock to 900mhz and then set for 1050mhz. Ran Furmark at Benchmark preset 1080 (1920X1080), then Benchmark (users settings 2560X1600 w/8x MSAA), then Benchmark preset 1080, then Benchmark preset 720 (1280X720). Both cards did the same thing. In 1080 and 720 mode the gpu osculated, but in 2560X1600 it stayed at a solid 99% gpu usage. Ran MSI AfterBurner v2.3.1 and GPUZ v0.6.7 Each of these were set to log.

[KEVIN_C 1/31/2013 7:32:12 PM] Hello we have actually seen that normally with powerful cards, usually with powerful benchmark settings, think of it as a vehicle rev limiter when the card gets to 100% it will fluctuate the clock speeds so as not to burn itself out. My question to you is does this ever happen in a normal game? It is not going to want to get to 100% and stay at 100% power so it tries to alleviate itself but fluctuating constantly when it gets that high.


So after all of this, I sent back the 7970's and got a couple of the EVGA 680 Classifieds. No Fluctuation in processor speeds now. As as I have seen others post about the 7950's doing the same thing. Have been a long time fan of the ATI cards since 1998.

Thanks for all the information, guys!

As to your questions: Yes, C3 does run on my SSD, I am just using a single card (which I was able to squeeze out a few extra MHz to 1180!), but I haven't checked the temperature yet. Will do now.

120+ FPS w/a 120Hz monitor is GLORIOUS.

There apparently quite a few XFX people talking about that issue and there is a fix. The solution was to adjust powercontrol settings +10%-20% in catalyst, which raises the max voltage the card can draw under load. Temps will go up a bit but it can prevent the clock from cycling under intensive loads eg 99% and 100% loads. Alternatively change to the non-boost bios and overclock the card a little bit manually.

Could be something to try. Tons of posts on XFX and Sapphire cards with boost and problems. Its a working as intended but not implemented properly problem with some boost cards.

Well, I do have the power control at 20% for precisely that reason. When the stutter happens, it's not even under full GPU load and it's sittin' pretty at 60 degree celsius. I may have to find a way to set the core clock = boost clock, or just disable it.

it happens because of the cpu, i have 2 systems with almost identical gpu's
GTX 550 ti OC and HD 7770 reference
the comp with the 550 ti stutters because of lower end processor(i5 2320) while the one with the 7770 is a i7 3770T at stock and i never get frame rate drops

Also, overclocking these cards gives me the DXGI device error
but, if i enable the over ride in console, error goes away

Well, I do have the power control at 20% for precisely that reason. When the stutter happens, it's not even under full GPU load and it's sittin' pretty at 60 degree celsius. I may have to find a way to set the core clock = boost clock, or just disable it.


Many of the boost cards have a second bios without boost for a reason, I would suggest switching to that XD. The problem with AMD boost is that it requires all the voltage and VRM to have a decent performance overhead. Though its still better than Nvidia boost where the alternative is just to crash. XD.


It reminds me of Intels first mobile processors with turbo boost. My i7 720QM was a beast processor but the turbo boost before sandybridge had a few issues. Its normal clock speed was 1.6ghz but it could boost one core up to 3.2ghz, and all four to 2.66ghz. The problem was that under normal circumstances clock speeds never reached the advertised turboboost. Except certain multithread tasks and programs caused it to glitch and run all cores at 3.2ghz and also not temp or speed throttle. >.> I was doing some encoding and I heard my laptops temp alarm so I thought it had hit 70oC. Nope I checked it and it was at 98oC... 99... 100... 103oC...

Nice bug to run into, it either does a full hardware stop or melts. The solution is to switch to the low power profile when it does that which is also sorta bugged and caps the CPU to 900mhz.

EDIT: I thought it may have been a HP problem, and it was XD. They implemented something wrong a heap of people complaining.