Sign in with
Sign up | Sign in
Your question

955 3.8 OC temps question

Last response: in Overclocking
Share
August 9, 2010 2:31:03 AM

I'm in the process of trying to reach a stable 3.8 OC and had a question or 2 regarding temps. After running Prime 95 for 15-20 minutes, my Core temps raise to 63-64 but my CPU temp stays in the mid-hi 50's. I know the AMD recommended max temp is 62, but I've heard differing opinions about whether that relates to the Core or CPU temp. If it's the CPU temps I should focus on, then my OC is in a lot better shape than if it's Core temps.

Also, considering I have a 212+ with push/pull and my ambient temps are pretty warm (@30 Celsius) are my temps OK? Idle is 40-42 cpu & core and Load is 57 cpu and 63-64 core. Finally a question regarding fans. On the Hyper 212+ I am using a CoolerMaster R4 Blue LED fan as push @ 1800 rpm but my pull is the Stock Hyper 212+ CoolerMaster fan at 1400rpm. Is the difference in RPM between the push and pull a major issue? I have a fan controller and I could lower the speed of my push to match the pull, but I figure higher rpm/cfm is better than having the 2 fans match each other. Any help to any of the above questions is sincerely appreciated.

Screen of GPU-Z and HW Monitor after 20 min. Prime 95:

More about : 955 temps question

a b à CPUs
a b K Overclocking
August 9, 2010 2:55:48 AM

In theory, having the stronger fan in pull will reduce "pooling" of warm air that may occur when the push is creating higher pressure than the pull fan. Both of my fans on my 212+ are 90CFm CM LEDs, and They were installed with the first installation of the HSF, so I don't have practical experience with different fan capabilities on this HSF (thus why I started with "In theory").

As for temps, I run my 965 c3 at 3.8ghz daily in a temperature controlled room. At 24c ambient, I see around 32c idle, up to 53ish under load. Your delta (difference between ambient and CPU temp) is very similar with slightly higher delta at load. I use the program CoreTemp to monitor my temperatures, which I believe is CPU, not Core readings. I haven't had a problem using this for my measurement, but I am also running in a cooler ambient environment. Also, I have gotten my 965 to rock stable at 1.41v, while you appear to be running at 1.43v. That difference in voltages may also be responsible for the higher load delta. Is this the lowest voltage you have been able to run stable at 3.8ghz?

I am currently using MX-2 for my Tim. You didn't mention what you are using, but if it is the stuff which came with the HSF, that could also be responsible for a few degrees difference in your load delta. The stock TIM is by no means horrible, but there are better ones out there that may shave a few degrees off.
m
0
l
August 9, 2010 3:01:19 AM

Thanks for the reply. The TIM is used was AS5 and it was over a week ago, so it's had time to cure. I actually set the VCore to 1.415 in BIOS but it jumps to 1.43-1.44 when under load (Prime 95). Not sure what I could do to get VCore to stay at one value.
m
0
l
a b à CPUs
a b K Overclocking
August 9, 2010 3:13:59 AM

You could try set it lower? Worst case scenario it will crash on boot, and you have to bump it back up. I have the opposite problem. I have it set at 1.425, but it doesn't go higher than 1.408, generally. It is only anecdotal with no real data to back it up, but it seems to me that the mid range Asus boards (anything below the ROG lines) seem to use a bit more voltage than the midrange 890 chipset Gigabyte boards like mine. Food for thought.

Our chips are very similar, with only a stock cpu-multiplier difference. Some believe that the 965s are higher binned (slightly more efficient) than the 955s of the same steppings, but the differences should be minimal. If anything, the inherent variance between chips (no two will perform precisely the same) might be the more likely explanation if you can't run a lower stable voltage.
m
0
l
!