PC performance loss after NVIDIA driver update

[Bacono10]

General decrease in PC performance

So yesterday I updated my NVIDIA drivers from GeForce Experience

(driver version 355.60)

and ever since that my PC's performance has dropped considerably.

Things are generally slower, but it is most most noticeable in games.

before the driver update I would get solid 60+ FPS in Battlefield 4 at Ultra settings, but now the average is 30-40 FPS on medium settings. Same situation for all my other games.

I have tried uninstalling the driver and reinstalling it, and installing an older driver to no avail.

any suggestions are appreciated

PC SPECS:
Strix GTX 970

AMD FX-8350

8GB RAM

600 WATT PSU



Nothing has been overclocked.

Thanks in advance.

 

[Bacono10]

I'm on version 22.2

 

[Bacono10]

I did some research on why my clock speed is only 1.4 GHZ, and found out it's due to "CPU throttling" which is weird because my CPU temps were pretty good last time I checked them, definitely not high enough to cause throttling.

Is there any way I could disable CPU throttling and make sure my temps stay low enough manually?

 

[Darkbreeze]

CPU throttling is a core feature of the system. It's there to protect against thermal issues. I already pointed out that it was throttling, the thing is, we need to figure out why.

I'd update to the latest bios to start with. I'd almost bet there is an issue with your motherboard or cpu.

 

[Bacono10]

I understand now. but isn't it supposed to ramp down my CPU when it gets too hot? not permanently underclock it?

It seems that no matter what I'm doing, or how hot it gets, it's always at 1.4 GHZ.

 

[Darkbreeze]

Yes, that's why I'm saying something is wrong. The system will downclock the cpu multiplier and reduce voltage, when it's not under a heavy or sustained load. When demand increases, it pretty much instantly returns it to full speed or at least a higher speed. Some fluctuation is normal. Being permanently at it's lowest possible clock speed is not, and that's what it's doing right now which is why your performance is terrible. It's like using a very old 1.4Ghz cpu with the rest of your hardware being high end. You have a severe bottleneck due to this.

We have to figure out why it's doing that. Try this, not as a permanent solution but as a temporary measure. Go into control panel, power options, click on "change plan settings", then "change advanced power settings", and then double click on "processor power management". Open "minimum processor state and set it to 100%, make sure the maximum processor state is also 100%. Check where it says system cooling policy and make sure it's set for active rather than passive. Reboot and see what it does, take another set of screenshots running prime95.

 

[Bacono10]

These were the settings I was already using when I took the last screeshot.

Is that a bad thing?

 

[Darkbreeze]

Those power settings were already at 100% on both minimum and maximum power, plus system cooling policy was set to active? Try turning off cool n quiet in the bios. Make sure that turbo core is enabled in the bios as well.

 

[Bacono10]

After double checking all of the settings, here are screenshot

 

[Darkbreeze]

Did you go into the BIOS and turn off Cool N Quiet?

 

[Bacono10]

Yes, Cool N Quiet is disabled.

 

[Darkbreeze]

Well, then I say you either have a faulty motherboard or cpu. I'm going to bring in another AMD expert to take a look at this thread in case I've overlooked something.

 

[Bacono10]

Ok.

I appreciate all the help and effort.

 

[Darkbreeze]

No problem, that's what we're here to try to do.

 

[damric]

There's a few red flags here.

1) The PSU quality is totally unsuitable for a build such as this. When you say fan facing up, do you mean that it is drawing in cold air from outside the case, or hot air from inside the case? If you must use a CX, be sure it is drawing in cold air.

2) The 970 Gaming has shown to have rather weak power delivery, especially when coupled with a poor PSU. Basically if your PSU isn't filtering properly it makes the motherboard work that much harder. Not a good combination for pushing an FX-8 core.

3) Try rolling back your drivers.

4) Make sure that your power settings in Windows are all set to maximum performance.

5) I'm concerned that your case has poor airflow. Try running with the side panels off.

Otherwise, darbreeze has already identified the problem, and that is your CPU is throttling. It could be due to the motherboard being overloaded as I stated above, but you also need to check to be sure that your CPU cooler is mounted properly and that the TIM application is adequate. The Hyper 212 style coolers still have nasty gaps between the direct touch heatpipes, requiring more TIM than usual. I don't think anything is actually defective here, just weak components and they are set up wrong.

 

[Bacono10]

The PSU fan is facing upwards in to the case. I assume that means it's drawing air from inside the case.

By "rolling back my drivers", do you mean motherboard drivers?


All windows settings have been on performance.

I have tried running with my side panels off, and even with an external fan pointing at the PC.

I checked the CPU cooler and it is seated properly.

At the time of installing the heatsink, I actually used a little too much TIM ( this was less than 3 months ago)

I may be wrong, but my CPU temps seem to be fine. I don't think that's the problem.

About the PSU, is there one you recommend that would be a better fit for my build? Preferably one that is reasonably priced.

 

[Darkbreeze]

Crap, I didn't even notice that he said that. Yes, that is wrong. Your bottom mounted PSU needs to have the intake vent facing the bottom, not the top. Fan should be out the back, vent pointed towards the floor to bring in clean cool air and not radiate heat into the case. The PSU cooling system should be entirely separate from the rest of the case unless it's a top mounted unit. Thanks Damric. I don't know why I didn't notice that.

That could push an already borderline unit past the point of thermal compliance.

My only question on the throttling is that his temps don't show as hot, but that may be because it's running SO hot, that it's throttling immediately, from boot up onward but that seems unlikely since it wasn't having the problem before and nothing was changed in regard to the hardware being removed or replaced at that time, so far as I know. Any other ideas as to what might cause the system to throttle for no reason. The only thing I've ever seen cause that is either a faulty bios rom on the motherboard or a case of cool n quiet being schizophrenic.

Since temperature is USUALLY the reason behind throttling, I'd probably do as Damric suggests though and double check the paste job and mounting of the CPU cooler. It's probably a good idea to repaste and use this as a guide:

http://www.tomshardware.com/faq/id-2520482/solving-temperature-issues-hyper-212-evo.html

I have seen backplates get loose or mounting screws pop out before a time or two as well, especially if the unit was moved at all.

Thanks again Damric, and if you think of any other possibilities, please do throw them out here.

 

[Bacono10]

Alright, I will re-situate my PSU to face downwards, and see if that helps.

If it doesn't help, I will re-apply the thermal paste and re-install the heatsink using your guide.

 

[Darkbreeze]

Ok, hope one of them works out. We're starting to run out of possibilities here. After you do all that, I'd like you to take a look at your bios settings again and make sure that the data reflected there indicates whether or not the CPU is actually set to run at 4Ghz. Double check to see that your settings are "sticking" when you change them as well, in the bios, I've seen dead CMOS batteries cause some crazy things to happen. Never seen it cause throttling but you never know I guess.

Make sure your fan speeds are being reported and see what the CPU temps are reporting as while in the bios.

 

[damric]

Some less likely possibilities:

1) Defective heat pipes on the Hyper 212. It happens.

2) BIOS problems, causing P-state issues. Update or rollback BIOS. Also try a CMOS Clear.

3) Some Nvidia driver conflict with Windows. Rollback Geforce diver.

If worst case scenario and none of our suggestions will work, I do have a software method to "force" the P-states to keep it from throttling. I'd rather not go that route until we have verified all thermal issues since that could easily cause damage.

 

[Bacono10]

I have updated the BIOS, flipped the power supply, and double checked my BIOS overclock settings and CPU temps.

here is a screenshot of my BIOS after the update

I have also tried other NVIDIA drivers.

If it was defective heat pipes on the heatsink wouldn't it be an issue from the start? not just all of the sudden?

and there is no visible CMOS battery on my board.

 

[Darkbreeze]

All boards have a cmos battery. Yours is here:

Remove it for a minute, then put it back. You'll probably need to enter the bios afterwards, reset your fan profiles and save settings, maybe. I'd probably just double check them.

 

[Bacono10]

I had to remove the GPU to access the CMOS battery, but after clearing it nothing has changed.

 

[Marka612]

Did you wait for the box to be completely cool before putting the battery back?
You should do it in this order:
Let computer iddle, so that temperature get as low as posible.
Turn it off.
Wait until the computer is at ambient temperature
Remove PSU cable
Remove battery
Wait arround 10minutes (no kidding, some mobo take that long to reset)
put MOBO battery
Put PSU cable
turn on
Check

This is like my third post, but i worked on a computer shop for quite some time :3
If that doesn't fix it, pbby some component has been damaged.

 

[Darkbreeze]

CMOS resetting does not take ten minutes on any system known to me, which is the majority of them. I seriously doubt ANY system. Residual power is depleted within seconds, or at least no more than a minute. Generally, probably as soon as the battery connection through the CMOS batter is broken in reality.

Never in my life have I heard of letting the system "cool" before removing the cmos battery. Not in school, in practice or from any other tech, ever.

 

[Marka612]

Yeah... 10 minutes may be so much, but I've seen some systems keeping energy after arround 30 seconds without the battery and if you say 10 minutes you make sure the MOBO will get a full minute without the battery.
If a sensor doesn't detect the temperature changes well, it could be because it either keeps the last temperature state when powered down or because it takes the first temperature when powering on, both are terrible designs, i have seen them both. Altough the first case should anyway be fixed when removing the battery, one never knows.