Sign in with
Sign up | Sign in
Your question

X4 955 BE temps are starting to get high and HOT

Last response: in CPUs
Share
April 19, 2010 1:46:03 AM

I built my pc with a 955 BE amd and got aftermarket cooler the hyper 212+ which got the temps for the cores to be ranging from 30-39C. But just the past few weeks the temps are now fluctuating from 35-49C which slows down the pc a little at times and also has automatically shuts down and restarts about 5times the past month. Why is that?
I had reapplied the paste still the same, case is 922 haf, fans are clean, psu is 550v dont know how to read CPUID voltages to determine if theres a problem. On bios its set at 1.35V.
its creating problems and 35-49C are only when browsing web and videos.
So what else could be the problem everything else seems to be running cool just the core cpu
a b à CPUs
April 19, 2010 12:25:07 PM

49C is no motive for slowing or shutting down but you can set the voltage to 1.25 and the Bios shutdown protection to 55.
m
0
l
a b à CPUs
April 19, 2010 1:51:40 PM

not hot. pretty good in fact. past 62 is considered dangerous.

pc shutting down could be with bad ram. run a memory test to see.

http://www.memtest86.com/
m
0
l
Related resources
April 19, 2010 3:14:32 PM

Are you getting a BSoD with that? Don't forget to disable the "Automatic Re-start on Shutdown" option in "Startup & Recovery Options" if you want to get a look at any diagnostic info before it goes splat.
m
0
l
a c 133 à CPUs
April 19, 2010 4:42:27 PM

I would check your PSU a faulty PSU would cause everything you explained
m
0
l
a b à CPUs
April 19, 2010 6:00:28 PM

stockg86 said:
I built my pc with a 955 BE amd and got aftermarket cooler the hyper 212+ which got the temps for the cores to be ranging from 30-39C. But just the past few weeks the temps are now fluctuating from 35-49C which slows down the pc a little at times and also has automatically shuts down and restarts about 5times the past month. Why is that?
I had reapplied the paste still the same, case is 922 haf, fans are clean, psu is 550v dont know how to read CPUID voltages to determine if theres a problem. On bios its set at 1.35V.
its creating problems and 35-49C are only when browsing web and videos.
So what else could be the problem everything else seems to be running cool just the core cpu


What kind of CPU loads are you giving it or are you saing the idle temps suddenly increased? I've had my 955BE MBOX OC'd to 3.6 Ghz for months (BIOS auto voltage 1.37) now with the same hyper 212+ cooler and idle temps are 30 Cthe first 15 minutes and 33-34C after hours of basic load, surfing, etc...

Do you have CnQ enabled in BIOS?




m
0
l
April 19, 2010 8:25:27 PM

CnQ is enabled and it fluctuates the first 20min goin to 40C then after an hour or so of surfing,etc.. It stays in temps of 42-49C.
What are the most effective ways to test RAM, PSU, and reapplication of thermal paste? Also how do u undervolt i cant seem to do that? Is it bad to undervolt in the short or long run?
m
0
l
April 19, 2010 11:09:15 PM

So what does CnQ do really? and am i suppose to have it on or off which is better and whats its pros/cons
m
0
l
April 21, 2010 10:11:20 PM

ok well now i dont have a real issue on high temps but i do still get a few random shutdowns and i just recently got my first BOSD. Could it be my power supply not having enough juice or something? I have a antec 550W continuous. components are 955 BE, 5770 sapphire 1GB, asus MB, hyper212+, samsung HD, asus dvd drive, etc.


here's a pic on cpuid i dont know how to calculate the voltage necessary so please help

m
0
l
April 21, 2010 10:38:47 PM

What RAM do you have? VIN1 is usually your RAM voltage, and it could be that your RAM is over/undervolted with the current settings. Also, use a program such as SpeedFan to double-check your voltages, your 3.3V rail is coming up as .56V in HWMonitor, which is hopefully just a misread by the program.
m
0
l
April 21, 2010 10:50:48 PM

My ram is G.SKILL Ripjaws Series 4GB (2 x 2GB) 240-Pin DDR3 SDRAM DDR3 1600 (PC3 12800) Desktop Memory Model F3-12800CL7D-4GBRM. also the 3.3V is still .56 on speedfan
how do you max out load for the cpu to its max 3.2ghz so i can see the voltage on cpuz ive only seen it go to 2.1ghz 1.355v. it idles on .982V

m
0
l
a b à CPUs
April 21, 2010 11:33:56 PM

Prime95 is a good CPU stress test. It will run your CPU at 100% to test stability and heat output. These are the temps that matter most for deciding if a CPU is safe.

Your light-load temps could have been raising due to dust/background programs/spyware/viruses as well.

Memtest86 is where I always start for random crashes.
m
0
l
April 21, 2010 11:54:22 PM

well i ran prime95 and heres the new data when its on full load out. is the voltage of power supply sufficient or not?

m
0
l
April 22, 2010 12:41:20 AM

47 Degrees for your cores while running prime95 is not extremely high. It is decent, considering that I think your cpus max temp given from amd is ~62 degrees C.

What confuses me about your post is your previous screen shot, before you asked about/used prime95. If I am reading your post correctly that screenshot is under normal load, and shows 47 degrees for your cores...BUT also shows your 4 cores all running at 100 percent.
What was going on with your computer in that previous message, when you asked how to stress all 4 cores? Yet all 4 cores ARE in fact running at 100%.
m
0
l
a b à CPUs
April 22, 2010 1:03:42 AM

Um, i notice something everyone, Look at the +12v rail in speed fan. it's only giving 4v.

Psu is going bad.
m
0
l
April 22, 2010 1:56:40 AM

^Yes, but 3.3V is also giving .56V. It could be a faulty PSU, although it could be a misread. Do you have a spare PSU with sufficient wattage that you can through in there to test?
m
0
l
April 22, 2010 3:46:16 AM

so what should the correct voltage be for those readings +3.3v and +12v or anything else? and i dont have any spare psu only old ones that are 300w. so what can i do to check if faulty psu or misreads?
m
0
l
April 22, 2010 3:51:40 AM

If you have onboard video, you can take out your GPU and use onboard, the 300W PSU will be enough if it has all of the correct connectors. Boot into BIOS first and check out the voltages in your PC Health option, the +3.3V, +5V, and +12V should all be close to 3.3V, 5V, and 12V respectively. A little variance is acceptable, my Corsair VX550 has voltages of 3.31V, 5.03V, and 12.10V on those three rails.

If the voltages are still messed up in BIOS (e.g. 4.8V on the +12V rail), then I would swap out the PSU and check to see if the spare one gives any better readings.
m
0
l
April 22, 2010 8:16:30 AM

my motherboard ( ASUS M4A79XTD EVO AM3 AMD 790X ATX AMD Motherboard ) doesnt have onboard video. I havent had any real problems like this for first 2months since i built the pc. is there any other way to test the psu if its faulty other than buying a new one and i dont wanna use my old spare 4yrs old 300W dont trust it.

I was also wondering if to make things simpler since it seems to be my psu should i just buy this psu tester to see if its my psu to determine whether i need a new as well.
http://www.newegg.com/Product/Product.aspx?Item=N82E168...
m
0
l
April 22, 2010 8:43:38 AM

You can use that tester, it will definitely tell you if the power supply is defective or not. If you have access to a multimeter, you can use that to measure the voltage that is being put out as well.

Before buying anything, I would definitely run Memtest86 for 8+ hours to see if you can dismiss it being a problem with your RAM.
m
0
l
a b à CPUs
April 22, 2010 9:01:20 AM

4 volts on the 12v rail? Ive never seen a functioning system with that low a reading. I wonder if its really correct.

Another possibility to consider is software. What is the history of this operating system? Fresh install? Migrated from another rig?

You can test this theory somewhat with a boot disk or a fresh install on another hard drive if you have one. But I would continue to look very seriously at those voltages as well.
m
0
l
April 22, 2010 10:55:02 AM

well currently im running windows 7 ultimate 64bit but i did however have 32bit before that and it has kept the program files of the 32bit locked and on hard drive as it continues to install new programs and softwares to that program files(32 bit) rather than the 64bit program files. Could that be creating an issue of somekind because ive tried before formatting the harddrive before installing 64bit but it continued to keep the 32bit folders.

Also know any good reliable site to dl memtest86 im using 4gb ram? and is running these voltages that bad for all my components
m
0
l
April 22, 2010 6:12:13 PM

http://www.memtest86.com/ You can create a bootable floppy, or a bootable CD using one of their ISO files. (you will need a burning program to create the CD).

When doing a fresh install of ultimate, make sure when you get to the screen where it has you allocate disk space that you delete EVERY partition, so that you full HDD is "Unallocated Space". Then create your desired partitions. This will format everything on your computer. Win7 has a Program Files (x86) folder, and a Program Files folder (the plain one is the 64-bit) even on a clean install. But nothing from the old OS will be copied over if you delete everything when installing Windows 7.
m
0
l
a b à CPUs
April 22, 2010 8:56:05 PM

That 4 volts is a misread, it would not even boot if it were that far off. It could even be the 5V rail you are seeing. Check what the BIOS says, as HWMonitor can only guess which sensor is what, as there is no standardization.
m
0
l
April 22, 2010 9:22:59 PM

ext64 was right i checked bios and +3.3v is at 3.458v and +12v is at 12.444 vcore is at 1.357. so im still wondering what else could cause or create rare shutdowns that happen say once out of 12
m
0
l
April 22, 2010 11:42:56 PM

stockg86 said:
ext64 was right i checked bios and +3.3v is at 3.458v and +12v is at 12.444 vcore is at 1.357. so im still wondering what else could cause or create rare shutdowns that happen say once out of 12

I guess you just decided to skip over that suggestion when I made it. :sarcastic: 

Anyway, have you ran Memtest yet? I would bet it being a memory problem. When it crashes and restarts, do you get a BSOD? If so, what is the error code it gives (such as STOP 0x00000024)?
m
0
l
April 23, 2010 2:38:44 AM

@Look: sry it seems i did skip it when u did suggest it and im goin to run the memtest tonight since i wrk late shift. But 8hr run? Also ive only gotten BSOD once and it didnt give me any error jus that windows was not shutdown properly supposedly.
m
0
l
April 23, 2010 3:22:16 AM

Haha it is fine. Memtest tests each sector of your RAM, so it might not reach the flawed sector until late in the test (if it is a memory problem). You could boot with half of your RAM (i.e. 1 stick), and if the error still occurs you could run Memtest on that stick alone. But generally overnight will be able to produce an error in your memory.
m
0
l
a b à CPUs
April 23, 2010 3:24:27 AM

Depends on how rock solid stable you want it. At least two or three full passes is a good preliminary test.
m
0
l
April 23, 2010 8:03:27 AM

just to confirm everything should i leave both sticks of ram during the test or one? which is faster yet accurate? also the one i dl asks which to boot a single cpu or multilple (experimental) or something else? im guessing its multiple and sorry for such noob questions
m
0
l
a b à CPUs
April 23, 2010 12:22:31 PM

Well, if you only have one stick in, you will have to test the other later. I usually do it first with both, and if it fails try individually.
m
0
l
April 23, 2010 1:39:38 PM

k so it boots the cd tho it asks me to choose single cpu or multiple? after choosing one nothing happens the test advances no where it says no cache or home address?
m
0
l
a b à CPUs
April 23, 2010 6:48:21 PM

Wait a minute, you have a Radeon 5750 or 5770. Change the drivers.
m
0
l
April 24, 2010 4:01:22 AM

@mosox i have 5770 are you implying that the gpu is causing these shutdowns and need to get updated drivers?

Also i have ran 3 tests of memtest86 and all with passing colors no errors at all. so what else could be causing these shutdowns? cant be a virus or trojan i got avg and its clean scan
m
0
l
a b à CPUs
April 24, 2010 9:56:51 AM

It must be the video card.
m
0
l
April 24, 2010 2:07:14 PM

whats the latest driver for SAPPHIRE HD5770 1GB GDDR5 PCIE (New Edition) for windows 7 64 bit ultimate?
m
0
l
a b à CPUs
April 24, 2010 2:17:42 PM

10.3 is the latest official, however 10.3a and 10.4 preview are newer.
m
0
l
!