• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

3930k Overclocking HELP with ASRock Extreme6/GB

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

iTeRRoRiz3

Registered
Joined
Jun 7, 2012
Location
Chicago
I figured the overclocking up to 4.5ghz on this motherboard was as easy as switching one setting on the motherboard from auto to 4.6ghz. I originally did that but was wondering why my speeds were dropping I then discovered speedstep and I took it off but then realized it is a waste and runs way too hot if you are just doing basic tasks. So then I switched back to the 4.6ghz setting profile on the BIOS but I can no longer run SMALL FFTs on Prime95 anymore because the CPU runs way too hot instantly after I start prime95 but it will run blend just fine. I am wondering if my temps are fine right now or not I have a HAF922 with a NOCTUA DH 14 that I installed after the first time because my original temps were not meeting my expectations. Should I be able to run Small FFTs for longer than 1 second? I was running it early with my first heatsink install and I had it at around high 60s low 70s with the 4.6ghz bios setting. I have not set any voltage or anything I am simply using the bios to do it because it seemed to work before but now I cannot get good temps anymore or run prime95 after one incident. Any help would be great
 
I just woke up from leaving it off for about 7 hours and getting instant 80 degress celsius+ from the Small FFT test. I just ran the smal fft test on the stock 32 multiplier for 10 minutes and it did fine staying at about 65 degrees celsius whereas the small fft at 45 multiplier would be instant 80s. I also just tried small fft test at 4.0 ghz and it got to 80 in about 15 seconds. I am thinking that a 6 core at a full load for an air cooler overclocked just might be too much but does that mean that it is not stable? I do not think all 6 cores will ever be at 100% load especially for what I am doing. I have the 3930k bc it was for 180 on intel retail I just want to OC just to OC and see what I can get to. If it cannot run small FFT at 4ghz does that mean the OC is unstable because I did leave speedstep on like I said. I just tried playing BF3 at 4.5 and 4ghz they were the same temps at about high 60s around 68 two cores were getting to 71 the max for these CPUs are around 66 correct? I dont know if I would be comfortable running BF3 at 4.5ghz with the temps being that high so any suggestions on if you think the temps are high or what. Thanks
 
Last edited:
I remounted with ****ty thermal paste I had until the MX-4 I ordered comes, I am actually doing better with this thermal paste for anyone that goes to best buy or knows the company Dynex thats the paste I am using because I had extra from my i5 build and it actually worked better running at 4.2ghz after running Small FFTs for a minute all cores reach 66 degrees but I also ****ed up this mount the most out of any I plan to do it real nice when the MX-4 comes any suggestions, one blob, two lines, the most way I see people doing it is one small blob in the middle of the cpu. Also what kind of clock speeds should I expect on air maximum I could not find any threads of anyones OC temps on air. Thanks
 
Sandy Bridge chips don't OC much higher on water than air. I'd say 4.5GHz is about average, anything higher is increasingly lucky.
 
Sandy Bridge chips don't OC much higher on water than air. I'd say 4.5GHz is about average, anything higher is increasingly lucky.

Yeah I wasn't expecting anything more than 4.5 but I should stay below 66 correct?
 
very informational thanks guys.
Sorry this was missed...



Stay below mid 80's when stress testing. That will show low/mid 70's for most activities. 66C is fine, but not close to the max you would want for 24/7.
 
I just got my MX-4 yesterday anyone have a recommended way of putting on of putting it on a blob or two lines I am trying the blob way right now I have the same temps as the dynex.
 
I use MX-4 and I've tried a couple ways, but in the end (simplest / less mess) pull the CPU, clean-off the old TIM with Isopropyl alcohol, blow-off dust/lint, then using a spreader (credit card / real spreader) make a very thin even layer (less = more /translucent). Clean-off edges with 'spreader' and optional with a q-tip & alcohol. Remount CPU/HSF.

/edit - I've tried the 'line' and 'lines' and all too often you'll have a corner with no paste.

http://valid.canardpc.com/show_oc.php?id=2320509
 
Last edited:
All depends on your vCore. On an H100 (medium) with a <1.35v ~55C~65C or less depending on ambient temp and setup (case restrictions etc). My 4.8 @ 1.4v is 65C~72C with 24C-26C ambient temp. This second my idle is 32C~34C with 24C ambient temp.

I don't have my H100 in it's 'ideal' config, the fans are underneath the rad pushing through the roof which has a screen, the fans are stock and variable connected to the H100 head unit. The reason is to reduce maintenance and DUST. Ideal is to pull cool air into the case which is a dust magnet.

/edit - I was wrong about my ambient temp it's 79F (26C) - the temp outside my home office is 75F (24C) ; the SB-E is a nice room heater ;)
 
Last edited:
Depends on...

1. Vcore
2. Cooler
2a. The coolers' mount
3. ambient temps...

Basically, we have no idea.
 
Tight is the key and -0.5C sadly does't prove much if anything -- ambient changes poof +1C or more. Though very interesting thread.
Just quick note, for the TIM experiments, not only is the ambient temperature controlled by a professional lab's climate control system, the measurements are also made as 'delta ambient' instead. during the tests, we detected ambient changed no more than +/- 0.3'C, so the ambient surrounding energy should be consistent enough to rule out any irregularities.

What should 24/7 temps be at 4ghz? When Im on BF3 or just idling

we should be using Delta Ambient to talk, cause around the global ambient temp very different, there's also a curve to be adjusted, as in cold areas, the delta ambient can actually be lower due to surrounding performances. what is your Vcore at currently?
 
Just quick note, for the TIM experiments, not only is the ambient temperature controlled by a professional lab's climate control system, the measurements are also made as 'delta ambient' instead. during the tests, we detected ambient changed no more than +/- 0.3'C, so the ambient surrounding energy should be consistent enough to rule out any irregularities.

You're kidding (LAB climate control) - if the thermostat is 10" from the CPU maybe. If you want to prove the data - time based graph with 30 second - 1 minute increments (CPU + Ambient) then we'll have something to talk about. A core temp not to mention 6 cores can vary 5C and core temps bounce up and down 1C~3C and never remain constant. Next how do you scientifically determine torque - "finger tight + 1/4 twist"?! Bottom-line, if I need a microscope to see a difference with a ±0.3C (0.6C) margin of error then any of those TIM's & Blocks are statically identical.

By education I'm a Chemist and I am not trying to engage you in debate, but you need to understand from 'my' standpoint I can see huge holes in your hypotheses and testing methodology.

By error, too many threads today, I assumed the OP had an H100 and not a Noctua NH-D14 though they are close ; see - http://www.legitreviews.com/article/1801/12/
 
Last edited:
Sorry I have been inactive, I never adjusted my vcore infact I am still a scrub I did not even see it in the bios. I have decided 3.5 is a happy medium for me I guess. It is an NOCTUA NH-D14 my temps right now with 3.5 when gaming do not go above 45 with full loads not above 55. Any suggestions, comments, or tips?
 
Back