Best CPU Stability Test Software(s) ? for i7 3930k ? need 24/7 overclock !

brandon888 · Nov 21, 2016

Best CPU Stability Test Software(s) ? for i7 3930k ? need 24/7 overclock !

plan to 4.4 GHZ 24/7 if its possible under 1.30v or 1.35v max .... started with 4.0 GHZ at 1.22V used, prime , aida 64 for 1-1 hours .... alsoe intel Intel(R) Extreme Tuning Utility , LinX , OCCT ...
no problem yet ....

what is the best software to test ? Some people tell that even after 24 Prime test they got blue screen after 1-2 mins in crysis 3 or battlefield 4

here my speces

M/B - ASUS P9X79
CPU - Intel i7-3930K Processor
GPU - ZOTAC GeForce GTX 1060 Mini 6GB
Cooler - DEEPCOOL CAPTAIN 120 CPU Liquid EX
RAM - G.SKILL Ares Series 4 x 4GB 2400Mhz
HDD - Seagate 2TB 6GB
SSD - Samsung 850 EVO - 250GB
PSU - Corsair Enthusiast Series 850-Watt
Case - CM Storm Enforcer

geofelt · Nov 21, 2016

I do not like prime95 or IBT for stress testing.
They use specialized instructions not usual with desktop code.
I would look for something like OCCT which uses a more normal mix of instructions.

brandon888 · Nov 21, 2016

geofelt :

intel Intel(R) Extreme Tuning Utility not IBT

IBT is trash i agree .. even crushed at very high and maximum settings even at stock on my system...

so use OCCT CPU stress and run it for 1 hour ? 4? 12 ? 24? and use just cpu stress or cpu linpack ?
Keep in mind i don't OC ram .... just use multipier and set to 4.4 and v core to 1.30

geofelt · Nov 21, 2016

When stress testing, I usually run only long enough for the temperatures to stabilize.
OCCT gives you a graph to show progress.
Under normal heavy load, it will not get that high.
I would imagine that an hour would be more than enough.
I just tried linpack, Results seem similar.
I do not know the technical differences.

One other test would be to run memtest86+
You should be able to complete a couple of full passes with NO errors.

brandon888 · Nov 21, 2016

you might be surprized but my cpu passed linx for 50 min no errors and 70C maX

at 4.4 GHZ 1.30 V core

then i tested with OCCT CPU + Lipmax ... after 20 min i got bluscreened ... strange but i got only 54C

geofelt · Nov 21, 2016

Unless temperature gets really hot, like approaching 100c. it is not the temperature that stops you.
It is the ability to run your selected multiplier at the selected vcore.

You could leave the voltages on auto and see what vcore you get as you raise the multiplier.

Not all chips will do 4.4. I so not have the statistics for your 3930K.
But, in general, 6 core processors will not oc as high as their 4 core counterparts.

I do think 1.3v is about the right limit for your gen cpu.
So, perhaps you should be testing at 4.3.

Vellinious · Nov 21, 2016

For heavy load testing, use either OCCT or AIDA64. For light load testing, use the stability test in RealBench.

brandon888 · Nov 21, 2016

geofelt :

I pass OCCT CPU stress one hour no problem but fail On OCCT CPU + Linpack stress test .... though it was ram problem but memory passed all mem88 tests ... hm strange ..... played also battleifeld 1 for 1 hour .. no crush or blue screen ... temp under 50 all the time with 4,4 ghz )

NOfail On OCCT CPU + Linpack even at stock stettings ... i repeat even on stock with 3.2 ghz and turbo boost 3.8

something is wrong here

and also about V core ... on mainboard autoverclock asus set BCLK to 125 and leaves mullty at 3.3 it gives me like 4.2 ghz and it.s on 1.3 vcore ) i think thin i can go 200 mhz higher ^^

NerdIT · Nov 21, 2016

brandon888 :

When you start messing around with the CPU strap/BCLK, you have to make sure your RAM frequency is compatible with it. Some dividers only work with certain BCLK frequencies - thats why the autoOC changes it, it's most likely changing your RAM too. That is probably why you got a RAM error.

I would turn off auto/XMP and keep the strap/BCLK at 100Mhz - and then just up the multiplier, stress, if stable then raise it again, and repeat. Do this until you start getting crashes or blue screens - and then tune it back down to the last stable multiplier.

Then if you want, once you know you have a stable OC, you can try and crank up the RAM. Although, there apparently is a known issue with certain RAM with your chip... --> http://www.overclock.net/t/1336658/i7-3930k-ram-limitations-issues

Although BIOS updates may have fixed that by now -always make sure BIOS is up to date.

NerdIT · Nov 21, 2016

Other than benchmarks -you won't see much real-world improvements from OC'ing RAM. I leave it at stock frequencies, and sometimes have had to even downclock it for stability reasons.

An example, on my current rig -i7-5820k, X99/DDR4- the chip cannot handle RAM frequencies above 2400Mhz without stability issues - the chip just doesn't like it. This is "apparently" going to be fixed with a BIOS update -but I'm still waiting.. >.>

brandon888 · Nov 21, 2016

NerdIT :

i just compared auto overlock voltage to mine ^^ i don't use auto OC !

i have 1.3 v core and 44 multi thats all ... ram is on 1600 mhz and stock voltage 1.5 ^^

Seems like i can't Only pass OCCT Limpmax test ,,, can handle occt cpu test , aide 64 for 1 hour , prime 1 hour , inter extreme tune and ect

NerdIT · Nov 21, 2016

brandon888 :

Have you run the PSU stress test? I believe the Linpack test puts an extra heavy load on the power supply -that may have changed OCCT has been out for a long time -but it wouldn't hurt to check!

brandon888 · Nov 21, 2016

NerdIT :

brandon888 :

Have you run the PSU stress test? I believe the Linpack test puts an extra heavy load on the power supply -that may have changed OCCT has been out for a long time -but it wouldn't hurt to check!

will try OCCT PSU Test , but as you see i have corsair 850 watt with bronze .. pretty decent .... and gtx 1060 needs tiny power ^^ BTW gpu is at stock clocks ....

i will try now and write results tomorrow ... ty for advice !

NerdIT · Nov 21, 2016

brandon888 :

NerdIT :

will try OCCT PSU Test , but as you see i have corsair 850 watt with bronze .. pretty decent .... and gtx 1060 needs tiny power ^^ BTW gpu is at stock clocks ....

i will try now and write results tomorrow ... ty for advice !

No problem! By the way, I use Aida64 and it's stability test a lot and have for a long while, and I have found it to be the most accurate and reliable of the lot. If you can run that for an hour with CPU/FPU/Cache/RAM checked - you should be good to go!!

brandon888 · Nov 22, 2016

Passed Aida64 for an hour with CPU/FPU/Cache/RAM checked , 4.4 GHZ (44X100 BCLK) with 1.30V core ... do i need more tests ? temps are great as you see 58C maX )

Should i lover Voltaje or leave it and go for more stress ?

NerdIT · Nov 22, 2016

brandon888 :

I'd say you are good to go! 1.3v is what a lot of people will manually/statically set the CPU core voltage to when OC'ing these chips - so you are right on the mark.!

Your temps are awesome and voltages are fine. I would stick with the 4.4Ghz. If you decide you want to try and squeeze a little more out of it (4.5 or whatever) and you start crashing I would manually set the voltage to 1.3v and see if that gets you stable - but I would not go any higher than 1.3v. And remember when you set the voltage manually - it constantly stays at that voltage 24/7 -which can help with stability, but will increase temps, power consumption, and possibly shorten the lifespan of the chip.

Honestly I would just stick with your current stable OC at 4.4Ghz as I said -because even if you were able to get it any faster you wouldn't notice much if any performance increases -other than benchmarks- so there is no need.

Nice job!

NerdIT · Nov 22, 2016

At this point is where I would maybe try overclocking the cache multiplier. The closer you can get your cache (still called NB for NorthBridge in CPU-Z) frequency to your CPU frequency, the better. I set the minimum and maximum to the same so that the cache always runs at the desired speed -instead of idling down.

I would start with small increments, just like with the CPU -do not just crank it up all the way to equal your CPU frequency..you could fry your rig! This will raise temps so just watch out. Just increase by 100Mhz increment, stress, if stable increase again, repeat until instability/crash, and then dial it back down. Just like when doing the CPU.

Luck with OC'ing the cache (called uncore in some BIOS) will vary drastically from chip to chip. I've seen some people get it up to 4.0Ghz and above -but I crash if I get mine anywhere higher than 3.3Ghz - almost 1Ghz less than my CPU..

This is optional obviously - you will notice some performance increase if you can get a good cache OC. The system and apps will be a bit more responsive, a bit more quick and "snappy".

brandon888 · Nov 23, 2016

NerdIT :

brandon888 :

I'd say you are good to go! 1.3v is what a lot of people will manually/statically set the CPU core voltage to when OC'ing these chips - so you are right on the mark.!

Your temps are awesome and voltages are fine. I would stick with the 4.4Ghz. If you decide you want to try and squeeze a little more out of it (4.5 or whatever) and you start crashing I would manually set the voltage to 1.3v and see if that gets you stable - but I would not go any higher than 1.3v. And remember when you set the voltage manually - it constantly stays at that voltage 24/7 -which can help with stability, but will increase temps, power consumption, and possibly shorten the lifespan of the chip.

Honestly I would just stick with your current stable OC at 4.4Ghz as I said -because even if you were able to get it any faster you wouldn't notice much if any performance increases -other than benchmarks- so there is no need.

Nice job!

No idea how to OC chase to be honest

so i will leave it alone

I went for 4.5 GHZ at 1.3 but crashed after 2-3 aida tests ... as i see aida 64 is best stability test so thank for advice !

when it passed 1 hour at 4.4 ghz 1.3 v it did't crashed in games

and was stable !

Darkbreeze · Nov 25, 2017

I realize this is an older thread, but since I've been gone for a while and only recently returned I thought it appropriate to address what I feel are conflicts of factual information in this thread, at least, depending on how you implement the criteria.

Nothing at all wrong with using Prime95 to test stability and thermal limits. It's about HOW you do it that matters. Versions of Prime95 OTHER THAN version 26.6 are not suitable for this purpose unless you are planning to run applications that extensively use the AVX instruction set.

Prime95 v26.6 is THE primarily accepted way to do the majority of baseline stability and thermal limit testing running the Small FFT option.

Prime95 version 26.6: http://windows-downloads-center.blogspot.com/2011/04/prime95-266.html

Further, as explained to me and many others on this forum by Computronix, who has far more experience with CPU architectures and testing procedures than 95% of the people you will ever meet, speak to or read about. He is also the author of the Intel temperature guide, found here:

http://www.tomshardware.com/forum/id-1800828/intel-temperature-guide.html

This pretty well sums things up and is equally relevant whether working with an Intel or an AMD system.

I can think of several reasons why x264 encoding or AVX / AVX2 / FMA3 apps won't work as a unilateral metric for thermal testing.

(1) A steady-state workload gives steady-state temperatures; encoding does not.

(2) Simplicity in methodology; most users would find encoding apps unfamiliar and cumbersome to accomplish a simple task.

(3) Most users such as gamers never run any apps which use AVX / FMA, so adaptive or manual voltage aside, it makes no sense to downgrade your overclock to accommodate those loads and temps.

(4) Standardization; Prime95 has been around since 1996; many users are familiar with it.

For the minority of users who routinely run AVX / FMA apps, then P95 v28.5 can be useful tweaking BIOS for thermal and stability testing.

regardless of architecture. P95 v26.6 works equally well across all platforms. Steady-state is the key. How can anyone extrapolate accurate Core temperatures from workloads that fluctuate like a bad day on the Stock Market?

I'm aware of 5 utilities with steady-state workloads. In order of load level they are:

(1) P95 v26.6 - Small FFT's
(2) HeavyLoad - Stress CPU
(3) FurMark - CPU Burner
(4) Intel Processor Diagnostic Tool - CPU Load
(5) AIDA64 - Tools - System Stability Test - Stress CPU

AIDA64's Stress CPU fails to load any overclocked / overvolted CPU to get anywhere TDP, and is therefore useless, except for giving naive users a sense of false security because their temps are so low.

HeavyLoad is the closest alternative. Temps and watts are within 3% of Small FFT's.

Best CPU Stability Test Software(s) ? for i7 3930k ? need 24/7 overclock !

Honorable

Titan

Honorable

Titan

Honorable

Titan

Honorable

Honorable

Distinguished

Distinguished

Honorable

Distinguished

Honorable

Distinguished

Honorable

Distinguished

Distinguished

Honorable

Retired Mod

Share this page