AMD FX 8350 vcore and frequency drops

Toggle sidebar Toggle sidebar
Status
Not open for further replies.
J

Jalen1

Honorable
Apr 16, 2013
13
0
10,510
Today, while testing my overclock for stability, I noticed something that didn't seem quite right. While under full load, my CPU's vcore and frequency have (seemingly) random drops. The vcore drops from 1.46v to around 1.32-1.35v while the frequency drops from 4.5 ghz to about 3.4 ghz.

I have no idea why this is happening, and I'm not totally sure it's a bad thing, but I'd like to know if anyone can help stop this from happening if it is actually detrimental to my system's performance.

Here's a picture of a 3 minute window where you can clearly see the vcore and frequency drops: picture
 
Solution
cmi86
Well I after about 6 beers and a few hours I came to find out that HWMonitor was displaying incorrect temperatures about 10c short of actual so after downloading the motherboard tuning utility i found my cpu was hitting mid 60's instead of mid 50's and thermal throttling it's self. I just gave up on stress testing and put faith in the fact I have gamed and multi tasked the snot out of this thing without fail and also the realistic outlook that using my system the way i do I'm never going to have my CPU at balls out 100% for long enough to get it to throttle anyways. Food for thought ?
Sort by date Sort by votes
J

Jalen1

Honorable
Apr 16, 2013
13
0
10,510
In prime 95 or OCCT, I max out at about 55 using a 212 EVO.

I tried setting LLC to the highest, and that caused temps to go above 60 in OCCT (didn't try p95), but that didn't solve the problem.
 
Upvote 0 Downvote
8350rocks

8350rocks

Distinguished
Mar 14, 2013
6,235
0
18,460


LLC should best set at highest for a high voltage OC like that...but your voltage is pretty high....how far are you OC'ed? Most people I have seen hit 5.0 at quite a bit less voltage with stability.
 
Upvote 0 Downvote
cmi86

cmi86

Distinguished
Sep 29, 2010
2,146
0
20,310
I have a similar issue on my FX-6300 clocked to 4.5Ghz 1.42 idle 1.35 load and somewhat higher temps than I would have expected. I have yet to resolve my issue but I have seen others resolve their issue by disabling APM (application power management) in the BIOS.
 
Upvote 0 Downvote
J

Jalen1

Honorable
Apr 16, 2013
13
0
10,510
I'm only OC'd to 4.5, but I was getting occassional errors in p95 at voltages a little lower than that. I'll probably go back and do more tweaking, though. I did put LLC on max, but the vcore and frequency drops were still happening.

Also, I haven't seen anything in my BIOS about APM, perhaps my mobo doesn't have an option for it.
 
Upvote 0 Downvote
cmi86

cmi86

Distinguished
Sep 29, 2010
2,146
0
20,310


i have heard you can disable it in overdrive if your board doesnt give you the option. its a chip feature not a board feature.
 
Upvote 0 Downvote
J

Jalen1

Honorable
Apr 16, 2013
13
0
10,510

Well, I disabled APM (or so I think), and also set LLC to the second highest setting after dropping the vcore to 1.39. Didn't help at all, as seen by this picture:
n7Y15Vy.png
 
Upvote 0 Downvote
J

Jalen1

Honorable
Apr 16, 2013
13
0
10,510
Yeah, I was on high performance. I didn't have the passive cooling policy enabled, so I did that, but still no luck. I did notice there were no drops prior to hitting 45 degrees, but after that they were as prevalent as ever.
 
Upvote 0 Downvote
8350rocks

8350rocks

Distinguished
Mar 14, 2013
6,235
0
18,460
Sounds like temperature throttling...you may want to improve your cooling. Perhaps your temp sensors are off somewhere, you may be hitting higher temps than indicated at which point your system may throttle you down.

What is your cooling setup?
 
Upvote 0 Downvote
J

Jalen1

Honorable
Apr 16, 2013
13
0
10,510
Case: Corsair 500r
Intake: 3 120mm (2 front, 1 bottom), 1 200mm (side)
Exhaust: 3 120mm (1 back, 2 top)
CPU Cooler: 212 EVO push/pull with corsair SP 120s
Thermal Paste: Arctic Silver 5

Edit: Dropped the frequency and vcore down a bit, still the problem persists. I also noticed again that no drops occurred before 45 degrees. Could it have something to do with HT or NB? I haven't tweaked those at all, their settings are all set to auto.
 
Upvote 0 Downvote
J

Jalen1

Honorable
Apr 16, 2013
13
0
10,510
Just set EVERYTHING back to stock (vcore, frequency, llc, etc.), and the problem still persists. I still didn't appear to get any drops before 45 degrees, but this time, the number of drops was much less (3 frequency drops in 2 minutes as opposed to ~10).
 
Upvote 0 Downvote
8350rocks

8350rocks

Distinguished
Mar 14, 2013
6,235
0
18,460


Hmm...that's interesting...I will have to do some reading and see what could be causing the throttling...hold that thought...let me do some research.

EDIT:

I just had a thought...what PSU are you running?

Voltage fluctuations from PSU could be causing your instability and it would explain why after removing the OC, the same issue persisted, just to a lesser degree...

You may have a PSU problem. Also, if your household electrical wiring is older, this could contribute to the issue. It sounds like a power problem somewhere...

It *could* be the motherboard...but I would check the PSU first...if your house is really old...you can test the wiring theory by taking your PC to a friend's house and testing it to see if the results are the same. If they are, it is internal power related most likely. If they are not similar...it's your wiring being old/inefficient.

Both can be easily fixed...I will explain how when we determine which is the problem...(or end up back at square 1...)

Try that out and tell me what happens.
 
Upvote 0 Downvote
J

Jalen1

Honorable
Apr 16, 2013
13
0
10,510
I'm running a corsair CX750M.

Also, I'm living (currently) in university housing that is only 3 years old, so I don't think that's the case unless they tried something sneaky with the wiring.
 
Upvote 0 Downvote
8350rocks

8350rocks

Distinguished
Mar 14, 2013
6,235
0
18,460


Have your PSU tested to see if it's giving consistent voltage...also, you might try going to Uninterruptable Power Supply to plug your PC into...because that will ensure that the current going to your PSU is regulated and will fluctuate less. A UPS will normalize your electricity supply and prevent spikes and sags.

EDIT:

Depending on where you live, it could just be the electrical grid in general...the UPS option is a solid plan regardless...but you might want to have your corsair checked. They're generally good...but...nothing is perfect. Also, how many watts does your system draw at peak? Try something like this if you want to monitor your power consumption/supply

http://www.p3international.com/products/special/P4400/P4400-CE.html

"The Kill-a-Watt"
 
Upvote 0 Downvote
J

Jalen1

Honorable
Apr 16, 2013
13
0
10,510
I have no idea how many watts my system draws at peak, and I would rather not spend any money on troubleshooting tools just to find that that isn't the problem.

How would you recommend testing my PSU to see if it's giving consistent voltage? Is there a specific program that does a quality job of testing them?
 
Upvote 0 Downvote
8350rocks

8350rocks

Distinguished
Mar 14, 2013
6,235
0
18,460


You'd have to go down to a PC repair shop and have it tested...many places will check it with an ohmeter for virtually nothing as it's not a difficult process. (If someone tries to tell you it takes more than 30 minutes or so...they're trying to game you...) Someone well qualified to do it likely could in 5-10 minutes.
 
Upvote 0 Downvote
J

Jalen1

Honorable
Apr 16, 2013
13
0
10,510
Well, I didn't open the case and direct a fan onto the VRMs, but at stock clock and voltages, there were no frequency drops. I can't remember if I have LLC on or off (I want to say it's set to auto). I guess I can try bumping up the clock in a while, but I'm not going to do it now as I have things that need to be done.
 
Upvote 0 Downvote
cmi86

cmi86

Distinguished
Sep 29, 2010
2,146
0
20,310
Well I after about 6 beers and a few hours I came to find out that HWMonitor was displaying incorrect temperatures about 10c short of actual so after downloading the motherboard tuning utility i found my cpu was hitting mid 60's instead of mid 50's and thermal throttling it's self. I just gave up on stress testing and put faith in the fact I have gamed and multi tasked the snot out of this thing without fail and also the realistic outlook that using my system the way i do I'm never going to have my CPU at balls out 100% for long enough to get it to throttle anyways. Food for thought ?
 
Upvote 0 Downvote
Solution
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom