7700k @5ghz+ stability testing. Questions for the people with success

[JNavy89GT]

Ok, I have my 7700k that I ran on an Asrock Z170 Fatality K4. Would run 4.8ghz at 1.325v. Running at 5ghz seemed ok, except for OCCT would would immediately fail like within seconds. Added voltage would help it reach a minute or so, but then it'd die too. Past 1.36-1.375 added voltage just made things worse. Aida64, Realbench, 3dmark, games seemed to work fine at 1.35v. Figured it could be the AVX ratio thing I'd need to adjust. Seemed as good of an excuse as any to take the time to upgrade my mobo(wanted the lights anyway ). Got the Asrock Z270 Fatality K6. Updated to latest bios. Have the AVX adjustment now. But tried 5ghz w/out the AVX downclock and I got another 7700k for the lottery odds to increase. Still immediately fail OCCT at 5ghz(voltages tried 1.3-1.385), but this 2nd cpu seems better as it will pass other benches at 4.8-5ghz at 1.3v. With AVX downclock OCCT will pass it seems. Well as much as I ran it for 10-20mins anyway, but again AIDA is passing at 5ghz 1.3v on this new cpu. Other games, benches so far or stress tests seem ok. Running an Artic 240 AIO cooler and temps with AIDA 5Ghz and 1.3v running 60-70's C in general. Can spike to high 70's or 82C is highest Realtemp lists. But again, in general about 60-70's full LOAD with AVX ON, with AIDA.

So my question becomes, is anyone passing OCCT with full AVX over 4.9ghz. TBH neither of my 7700k's seem to want to do this in either Asrock board. Maybe their bios's aren't mature enough for >4.8ghz and AVX? Maybe OCCT is just buggy past that? Using latest version of OCCT btw; I got the "old" 7700k and Z170 K4 board to pass 1hr of OCCT old version, but next day I ran it and only passed 5-10mins or so before fail. New version seems to fail faster. Should I not care about the OCCT errors since other programs seem fine? Should I just settle for 4.8ghz since that seems to pass everything. I don't really want to use the AVX ratio as despite my eagerness to upgrade this board, the feature seems buggy as when I put the AVX to -2 and run 5ghz, It will indeed downclock to 4.8ghz with OCCT or AIDA, but it will also seemingly flip between 4.8ghz and 5ghz just sitting and messing about in windows or browsing the internet, and that's not right to me :/. Seems it should hold a solid 5ghz until true AVX needs and thermal limits exceeded etc.... So to me Asrock has work to do with the AVX ratio hold. Or maybe that's how it is on other boards too and I'm just expecting too much?

Wanted some imput as I'm kinda OCD'ing myself to death here

 

[Version_3]

7700K here, with a Asus Prime 270-AR.

I was never able to pass anything with full AVX above 4.8GHZ in any stress/benchmark tool, including OCCT. Temps got way too high to even be feasible to run full time. I would see it hit TMAX consistently. Just to get my second core to not crash immediately upon start in P95, I had to up the voltage over 1.4v

I was able to pass OCCT at 5GHz on my 7700K with AVX downclocked to 4.7GHz. My VCore is set to 1.3v. Temps stay in the low to mid 70's while executing.

However, I got some pretty bad blue screens and crashes in games, so I backed down to 4.9GHZ (AVX downclocked to 4.7 still), and I've been 100% stable since.

 

[JNavy89GT]

ok thx mate.

 

[TordanGow]

My 7700K seems to hit a wall at 4.8GHz, at least with my cooling and no delid. I am 8 hour prime with AVX stable at 4.8 GHz with 1.24 vCore. Also 8 hour RaelBench stable. Temps on Prime would hit mid 80s regularly. I can boot into windows at 5 GHz @ 1.31 vCore and start testing, but with no delid I hit the thermal wall and throttle even on my NH D14.

Asus Z270a-Prime board BTW.

 

[Gideon]

A very few 7700K will pass a stress test at 5 ghz with AVX enabled. Most people running at 5 ghz or more are disabling it. Also delidding helps as running with AVX on makes the temps skyrocket.

 

[JNavy89GT]

I can pass AVX with Aida64 but not occt :/. Go figure. Realbench is stable. Think I'm gonna call it good at 5ghz here and just leave it be lol. Well, except for .............. hehe!

 

[Ocellaris]

Run the newest version of Prime 95 on Small FFTs and see how it does.

 

[SvenBent]

Aida64, Realbench, 3dmark are all "weak" CPU test they will happily runs for hours on an unstable system thta cant handle a few min of Linpack or Prime.
Linpack and PRime95 are a class of its own when it comes to CPU stress level.

What suits you is up to you, but i am not supriced an unstable system, runs the 3 weak test and not the real stuff.

 

[Slade]

AVX limit is 4.9 for me. 1.34V. Delidded and the temps can still hit 90s with AVX loads (Prime 95 is brutal with small FFT). I ran 20+ hours at those temps to be sure. I was so close to AVX 5.0. I think I noted that OCCT doesn't hammer the CPU as hard as p95 does with regards to AVX load. If you run p95 small fft, it pegs the temps for AVX, while OCCT seems to only partially use AVX in its testing, temps are sporatic with OCCT 70s average.

I think if you want to quickly check AVX stability, run p95 small fft. It will fail out or run high temps fairly quickly, I think OCCT and Realbench will pass if your CPU handles p95 AVX small fft for extended duration.

 

[Maxiom]

AVX limit is 4.9 for me. 1.34V. Delidded and the temps can still hit 90s with AVX loads (Prime 95 is brutal with small FFT)

Delidded + water and still 90s
I hope those instructions are worth the power they consume
What's your ambient temp btw?

 

[Araxie]

Aida64, Realbench, 3dmark are all "weak" CPU test they will happily runs for hours on an unstable system thta cant handle a few min of Linpack or Prime.
Linpack and PRime95 are a class of its own when it comes to CPU stress level.

What suits you is up to you, but i am not supriced an unstable system, runs the 3 weak test and not the real stuff.

You are actually partially wrong, what really matter is real world, you can have a successful stress test pass and still as soon as you load any realworld application, being work or game, or whatever and it will crash, specially when using offset and adaptive voltages, as different loads scenarios put different amount of vcore load and with the addition of vdroop or LLC things can be pretty stable under a full load stress test as the vcore remain constant and almost fixed for the entire length of the test, while real world applications and games put different loads and can insta-crash if the settings aren't tweaked properly,

Lot of my systems can pass any kind of synth stress test and still fail while gaming. A real Stable system have to be tested under both scenarios unrealistic workloads and realworld usage and that's what I found the bets on RealBench not in the stress test but in the benchmark with infinite loops. I can pass 24hours of the stress test and still fail in the first loop of the benchmark, that's why I always recommend the usage of 40+ loops of benchmarks in realbench instead of the stress test, specially for the guys who like to use adaptive and/or offset instead of fixed voltages.

 

[SvenBent]

You are actually partially wrong, what really matter is real world, you can have a successful stress test pass and still as soon as you load any realworld application, being work or game, or whatever and it will crash, specially when using offset and adaptive voltages, as different loads scenarios put different amount of vcore load and with the addition of vdroop or LLC things can be pretty stable under a full load stress test as the vcore remain constant and almost fixed for the entire length of the test, while real world applications and games put different loads and can insta-crash if the settings aren't tweaked properly,

Lot of my systems can pass any kind of synth stress test and still fail while gaming. A real Stable system have to be tested under both scenarios unrealistic workloads and realworld usage and that's what I found the bets on RealBench not in the stress test but in the benchmark with infinite loops. I can pass 24hours of the stress test and still fail in the first loop of the benchmark, that's why I always recommend the usage of 40+ loops of benchmarks in realbench instead of the stress test, specially for the guys who like to use adaptive and/or offset instead of fixed voltages.

Show my a test where prime95 cleans and you "Real world test failed and I'll show you a person in 99% of the cases that don't understand the difference between CPU testing and system testing.
Sadly that is way to common and people quickly mistake it for a failure on prime95 rather than people not knowing when to use a hammer or screw driver

As you said yourself you instant crashes due to a different load. its not the software fault the person cant figures out the right tool to establish the load for you.
if you clear prime95 and then crashes in games the issues is probably not the CPU but more likely your PSU or Motherboard, (graphcis cards) that just don't like the added stress/heat/powerload

And again as i wrote they are weak CPU tests i didn't say system tests. There is a difference

 

[Araxie]

Show my a test where prime95 cleans and you "Real world test failed and I'll show you a person in 99% of the cases that don't understand the difference between CPU testing and system testing.
Sadly that is way to common and people quickly mistake it for a failure on prime95 rather than people not knowing when to use a hammer or screw driver

As you said yourself you instant crashes due to a different load. its not the software fault the person cant figures out the right tool to establish the load for you.
if you clear prime95 and then crashes in games the issues is probably not the CPU but more likely your PSU or Motherboard, (graphcis cards) that just don't like the added stress/heat/powerload

And again as i wrote they are weak CPU tests i didn't say system tests. There is a difference

if you have used RealBench you should know that the "benchmark" contain several type of loads, 2 of those are Strictly CPU only intensive task and one of those use heavy AVX load with handbrake h.264 video compression however so it will also use other CPU Resources that P95 doesn't use, the same as the GIMP image editing part which can use SSE4.2 and AVX as single, or SSE4.2 only for the "Heavy Multitasking" section which use both Handbrake and GIMP at the same time will be better test than p95 or the newer linpack.

im not involving a 3rd person in any of my test, just about with all of my experience in any of my 14 active Machines which 10 of those are heavily overclocked both intel and AMD, im not a new overclocker and I certainly know the difference between CPU testing and system testing and that's why precisely im telling you that you are partially wrong, I work with heavy VM, Remote Client access and data bases environment for sure I know pretty damn well how to keep my machines stable under any kind of loads, im telling you all depend on motherboard choices and CPU voltage methods, for guys who use fixed/static voltages in BIOS yes everything what you say is good and apply without any question, for others who like to use offset methods and Adaptive methods to allow the vcore fluctuate between idle and stress, the "stability" word should be taken with a lot of more care as it can easily have a fully load and stressed CPU but as soon as it face games as Crysis 3 ie, it will crash or present issues after couple of minutes or even instant as the voltage isn't fixed, is constantly fluctuating. So to ensure fully stability real-world scenarios is the best to use and I said above in those strictly cases, I don't know what kind of experience you have with overclocking modern systems, but trust me is a whole different thing than what it was in the past, lot of people now uses alternative CPU vcore methods which require take stability precautions even on IDLE. is not common to hear/read nowadays "system fully stable on load, but lock-up while doing nothing, and/or sleep/hibernate" and for those p95 or linpack can't do any shit as those are mostly using offset voltages which are applied different depending on load.

 

[SvenBent]

if you have used RealBench you should know that the "benchmark" contain several type of loads, 2 of those are Strictly CPU only intensive task and one of those use heavy AVX load with handbrake h.264 video compression however so it will also use other CPU Resources that P95 doesn't use, the same as the GIMP image editing part which can use SSE4.2 and AVX as single, or SSE4.2 only for the "Heavy Multitasking" section which use both Handbrake and GIMP at the same time will be better test than p95 or the newer linpack.

im not involving a 3rd person in any of my test, just about with all of my experience in any of my 14 active Machines which 10 of those are heavily overclocked both intel and AMD, im not a new overclocker and I certainly know the difference between CPU testing and system testing and that's why precisely im telling you that you are partially wrong, I work with heavy VM, Remote Client access and data bases environment for sure I know pretty damn well how to keep my machines stable under any kind of loads, im telling you all depend on motherboard choices and CPU voltage methods, for guys who use fixed/static voltages in BIOS yes everything what you say is good and apply without any question, for others who like to use offset methods and Adaptive methods to allow the vcore fluctuate between idle and stress, the "stability" word should be taken with a lot of more care as it can easily have a fully load and stressed CPU but as soon as it face games as Crysis 3 ie, it will crash or present issues after couple of minutes or even instant as the voltage isn't fixed, is constantly fluctuating. So to ensure fully stability real-world scenarios is the best to use and I said above in those strictly cases, I don't know what kind of experience you have with overclocking modern systems, but trust me is a whole different thing than what it was in the past, lot of people now uses alternative CPU vcore methods which require take stability precautions even on IDLE. is not common to hear/read nowadays "system fully stable on load, but lock-up while doing nothing, and/or sleep/hibernate" and for those p95 or linpack can't do any shit as those are mostly using offset voltages which are applied different depending on load.

I know realbenchi i also know it uzes 7-zip which is a very low level load on the CPU and mostly integeras calculations so the load and heat developing is not there
7-zip is even a lighter load then simple Deflate optimizing with pngout.

Handbrake x264 is also lower load than prime 95 and Linpack.

Calling them intensive is a pure matter of subjectiviness, but they are objectively less a burden on the CPU than prime95 and Linpack

Saying that you don't involving any 3D in a turth with modifciation if you are gunning for realtemp since it uses a OpenCL load to be put on the CPU thta is why realbench is a system stability test and not just a CPU stability test.
I'm actually not really sure where you are going with this. first you are arguign that you aren o using 3d to stabilty test you things and then you are talkign about how it will crash in crysis. you are just making my point here
there is a diffren between a CPU stress test and a system stress test. if you machine erros out in crysis but not in prime95 it not becuase prime95 is weaker to stress out your CPU. its because crysis has a system wide load that prime95 does not.
that does snot mean prime95 can't be harder on the CPU.

Do you have anything more thna anecdotalc evdien that PRime95 and Linpack are not classe better than Aida64 REalbench or 3dmark in CPU testing ?
Pleae stay on the goal post here. You "Corrected" me in my statemnt but you are more and more moving away from the statmens to say something that is not what i said.


I said
- Aida64, Realbench, 3dmark are all "weak" CPU test they will happily runs for hours on an unstable system that cant handle a few min of Linpack or Prime.
- Linpack and PRime95 are a class of its own when it comes to CPU stress level

Whatever you system crashed due to crysis is irrelevant for thisdebat between us because as we both states its not just a CPU load with crysis.

 

[SvenBent]

just to bring in some actual evidence

Core I7 2700K		4.4GHz +0.030v			4.4GHz +0.20v			4.5GHz +0.20v			4.6GHz +0.20v			4.7 +0.20v
			Status	temp	Power		Status	temp	Power		Status	temp	Power		Status	temp	Power
Asus RealBench		OK	79-84	318w*		OK	80-85	318w		OK	80-87	303-319/326	C  0:22	76-83	xxx/325		C 0:00	XX-XX	xxx/315
Aida64 Extreme		OK	74-81	181w		OK	73-79	176-180w	OK 7:40	74-81	180-185/187	OK 4:50 76-82	182-186/190	C 0:15	73-82	xxx/193

Core Damage 0.8h					OK	76-82	180w		OK15:28	79-85	184/188		OK 9:50	79-85	186/190		C 1:16	79-85	xxx/192
Cpu Burn-in 1.01									OK 4:05 73-77	181/184		OK 8:15	81-77	185/187		C 0:06	73-78	xxx/190

Folding@Home		OK	73-78	182w		OK	73-79	179w		OK 9:30	73-79	180/182		C  2:49	74-81	xxx/186		C 0:00	31-31	xxx/185

HeavyLoad		OK	72-76	184w		OK	72-76	182w		OK 4:20	73-78	187/189		c  2:59	74-80	xxx/193		C 0:00	xx-xx	xxx/193
Handbrake 1.0.2		OK	71-76	165-177w	OK	72-76	96-180		OK 8:08	73-77	95-180/185	OK 6:30	73-80	160-182/186	C 0:29	72-76	xxx/186

IBT 2.54 Standard					OK	78-88	167-211w	OK 4:00	80-90	171-215/216	C  0:02 71-79	xxx/218		C 0:00	xx-xx	xxx/215
IBT 2.54 Maximum					OK	80-91	106-212w	C  1:19	83-94	xxx/220		C  0:02	72-82	xxx/217		C 0:01	63-68	xxx/206

LinPack V11.2.2.010	OK	78-90	191-210w	OK	78-89	162-207w	OK 4:15	80-91	161-211/213	c  0:05 78-88	xxx/218		C 0:00	xx-xx	xxx/210
LinPack V11.2.2.010 x2	OK	84-96	220w		OK	81-92	169-216w	C  0:55	82-94	xxx/223		c  0:01	73-76	xxx/223		C 0:00	xx-xx	xxx/182

OCCT 4.4.3 (Own/large)	OK	78-88	211w		OK	76-84	185-204w	OK 4:00	77-87	198-205/211	C  0:06	73-81	XXX/212		C 0:00	31-31	xxx/193
OCCT 4.4.3 (Linpack90%) OK	82-93	207-214						OK 4:05 74-83	191-196/199	C  0:48	76-85	xxx/204		C 0:02	37-53	xxx/199

Prime95 26.6 (small)									OK 4:05	77-85	195/198		C  0:12	76-85	xxx/201		C 0:00	xx-xx	xxx/195

PRime95 28.5 (8-8-30)	OK	84-93	217w						C 0:40	82-93	xxx/218 	C  0:00 68-76	xxx/215		C 0:00	xx-xx	xxx/213
Prime95 28.5 (small)	OK	79-89	212w		OK	80-90	211w		C 1:52		xxx/219		C  0:00 63-72	xxx/214		C 0:00	xx-xx	xxx/215
Prime95 28.5 (Blend	OK	82-92	215w		OK	84-94	201w		C 2:26	83-93	xxx/226		C  0:03	67-71	XXX/196		C 0:00	36-36	xxx/194

Passmark BurnInTest 8.1	OK	67-72	104-180		OK	67-72	103-178w	OK	69-73	105-179w/183	OK 6:35	69-74	108-175/182	OK 4:25 70-74	109-183/187
XTU (CPU)		OK	73-79	176-190w					OK 7:25	73-79	173-186/188	C  2:17	73-81	xxx/210		C  0:04	67-73	xxx/191

Winrar			OK	60-67	151w		OK	60-67	150w		OK	61-67	15-/155/177	OK11:00	62-69	155/156		C  1:20 61-67	xxx/165
7-zip			OK	68-71	124-173		OK	68-72	126-171w	OK 9:20	70-74	126-175/178	OK 9:40	71-74	128-177/179	OK 6:40	71-76	134-179/181


* This uses OpenCL to put load on GPU as well.
Not direct comparable against other candidates

Status key
C = Crash/BSOD
R = Reboot/Reset
D = Error detected
[code]

 

[Ocellaris]

just to bring in some actual evidence

Core I7 2700K        4.4GHz +0.030v            4.4GHz +0.20v            4.5GHz +0.20v            4.6GHz +0.20v            4.7 +0.20v
            Status    temp    Power        Status    temp    Power        Status    temp    Power        Status    temp    Power
Asus RealBench        OK    79-84    318w*        OK    80-85    318w        OK    80-87    303-319/326    C  0:22    76-83    xxx/325        C 0:00    XX-XX    xxx/315
Aida64 Extreme        OK    74-81    181w        OK    73-79    176-180w    OK 7:40    74-81    180-185/187    OK 4:50 76-82    182-186/190    C 0:15    73-82    xxx/193

Core Damage 0.8h                    OK    76-82    180w        OK15:28    79-85    184/188        OK 9:50    79-85    186/190        C 1:16    79-85    xxx/192
Cpu Burn-in 1.01                                    OK 4:05 73-77    181/184        OK 8:15    81-77    185/187        C 0:06    73-78    xxx/190

Folding@Home        OK    73-78    182w        OK    73-79    179w        OK 9:30    73-79    180/182        C  2:49    74-81    xxx/186        C 0:00    31-31    xxx/185

HeavyLoad        OK    72-76    184w        OK    72-76    182w        OK 4:20    73-78    187/189        c  2:59    74-80    xxx/193        C 0:00    xx-xx    xxx/193
Handbrake 1.0.2        OK    71-76    165-177w    OK    72-76    96-180        OK 8:08    73-77    95-180/185    OK 6:30    73-80    160-182/186    C 0:29    72-76    xxx/186

IBT 2.54 Standard                    OK    78-88    167-211w    OK 4:00    80-90    171-215/216    C  0:02 71-79    xxx/218        C 0:00    xx-xx    xxx/215
IBT 2.54 Maximum                    OK    80-91    106-212w    C  1:19    83-94    xxx/220        C  0:02    72-82    xxx/217        C 0:01    63-68    xxx/206

LinPack V11.2.2.010    OK    78-90    191-210w    OK    78-89    162-207w    OK 4:15    80-91    161-211/213    c  0:05 78-88    xxx/218        C 0:00    xx-xx    xxx/210
LinPack V11.2.2.010 x2    OK    84-96    220w        OK    81-92    169-216w    C  0:55    82-94    xxx/223        c  0:01    73-76    xxx/223        C 0:00    xx-xx    xxx/182

OCCT 4.4.3 (Own/large)    OK    78-88    211w        OK    76-84    185-204w    OK 4:00    77-87    198-205/211    C  0:06    73-81    XXX/212        C 0:00    31-31    xxx/193
OCCT 4.4.3 (Linpack90%) OK    82-93    207-214                        OK 4:05 74-83    191-196/199    C  0:48    76-85    xxx/204        C 0:02    37-53    xxx/199

Prime95 26.6 (small)                                    OK 4:05    77-85    195/198        C  0:12    76-85    xxx/201        C 0:00    xx-xx    xxx/195

PRime95 28.5 (8-8-30)    OK    84-93    217w                        C 0:40    82-93    xxx/218     C  0:00 68-76    xxx/215        C 0:00    xx-xx    xxx/213
Prime95 28.5 (small)    OK    79-89    212w        OK    80-90    211w        C 1:52        xxx/219        C  0:00 63-72    xxx/214        C 0:00    xx-xx    xxx/215
Prime95 28.5 (Blend    OK    82-92    215w        OK    84-94    201w        C 2:26    83-93    xxx/226        C  0:03    67-71    XXX/196        C 0:00    36-36    xxx/194

Passmark BurnInTest 8.1    OK    67-72    104-180        OK    67-72    103-178w    OK    69-73    105-179w/183    OK 6:35    69-74    108-175/182    OK 4:25 70-74    109-183/187
XTU (CPU)        OK    73-79    176-190w                    OK 7:25    73-79    173-186/188    C  2:17    73-81    xxx/210        C  0:04    67-73    xxx/191

Winrar            OK    60-67    151w        OK    60-67    150w        OK    61-67    15-/155/177    OK11:00    62-69    155/156        C  1:20 61-67    xxx/165
7-zip            OK    68-71    124-173        OK    68-72    126-171w    OK 9:20    70-74    126-175/178    OK 9:40    71-74    128-177/179    OK 6:40    71-76    134-179/181


* This uses OpenCL to put load on GPU as well.
Not direct comparable against other candidates

Status key
C = Crash/BSOD
R = Reboot/Reset
D = Error detected
[code]

That's the old version of Prime 95 without AVX.

 

[SvenBent]

What you see above is REAL testing of stress software not just you anecdotal sh#t

Asus real bench give A lot less heat and runs a lot longer on CPU unstable systems than prime95 and Linpack does
HOWEVER you will notice that the power usage is a lot higher because Realbench is a SYSTEM stress test and not just A CPU stress test.
But looking clearly at the CPU part of it, it is NOT near as intensive as prime95 and Linpack

You will also notice how Aida 64 runs for more than 7hours and 40mints (Before i stopped it) on a system that crashes at 40 min in prime95
Looking at handbrake we again see the same behavior where handbrakes runs for 8hours no problem (Again i stopped it so the real time could be a lot longer) in a system that crashes in 40mins in prime95
You can also see that aida64 runs for 4hours on a system that crashed within 5mins of Linpack testing.

The power load is higher on prime 95 and Linpack that it is on Aida64 and Handbrake and 7-zip (last 2 which is a part of Realbench,
The CPU temp is a lot higher on prime95 and Linpack than it is on aida64 and Realbench etc etc

So in short: Prime95 and linpack.
- Pulls more power
- more sensitive to instability
- generates more heat

So yes they are in any way,shape and form , a much more CPU intensive test than realbench handbrake 7-zip aida64 etc.

It was true in this test
It was true in the test i did 4 years ago as well.
This has been true in core2duo system i7 system and AMD system as well.


I'm i saying that you cant have shit crash with realbench and not prime95? NO because Prime95 is a CPU test and Realbench is a SYSTEM test
So again I'm not surprised by OP results as they are EXACTLY the same as we see in in a proper documented testing environment



P.S.:

Also please note this was test with a dynamic voltage with offset voltage. Not a set voltage.

The real jokes in this is passmark burnin test that never crashed within 4 hours ( the cut off time for a pass)
and 7-zip which again is part of the "intensive CPU" test of realbech. of all the ~20 test. 25% of the CPU load from Realbench is from the WORST CPU test candidate

 

[SvenBent]

That's the old version of Prime 95 without AVX.

Read again. I tested both 26.6 (no avx) AND 28.5 ( with AVX)

The only thing i kinda wished was that i had tested Prime in large mode as well and more of the 26.6 version.
But I already had plenty of prime95 trials in there and I would rather want to add in more software instead.

 

[SvenBent]

oh also another funny thing i saw.

The test was based on 4hours of stress testin.g if a software pass the CPU after 4 hours it would be an ok.
but due to the facts some software dont have a timer and i had to run some test while i was out. they got to run longer than that..
they would stil lhave recievde an OK if they would have run 4 hours and then crash on the 5th.

HOWEVER none of the long time running software crashed after 4 hours.
So far all the test that ran for more than 4hours ran until i stopped which could indicate that there is s strong diminishing returns on stress test more than around 5 hours.
However more testing would bee needed to verify that. Especially for the stronger canditates (priemn95 and Linacpk)

 

[Slade]

My non load ambient temps are in the low 30's. Under gaming loads, I see 50's. Delidding works. If it can be prime95 AVX stable for a day's worth of 90C then it will survive my more mundane day to day life. I put this thing thru 12 hrs of p95, 8 hours of realbench and 24 of occt. I followed that up with another 12+hr of p95. Stability testing to me is to show that at its worst, its stable, and since I will never put it to its worst in my normal usage, everything is great.

So in a word, stable is passing ALL of that and being able to game to your heart's content. If you are crashing in anything, then you have a problem. My previous x58 build ran 4.0ghz for 7 years 365 days a year. I stressed it HARD at the beginning, p95, orthos, intel burn test, all running at the same time for 48+ hours. It passed and I relied on this rig from Jan 2010, to Feb 2017.

For the hell of it, I've been running p95 all day since I first posted, I'm hovering under 90's now, 87-88 on core1, while the others are dangling around 85C It looks like my thermal paste has settled in since I first built this system a few weeks ago.

 

[SvenBent]

My non load ambient temps are in the low 30's. Under gaming loads, I see 50's. Delidding works. If it can be prime95 AVX stable for a day's worth of 90C then it will survive my more mundane day to day life. I put this thing thru 12 hrs of p95, 8 hours of realbench and 24 of occt. I followed that up with another 12+hr of p95. Stability testing to me is to show that at its worst, its stable, and since I will never put it to its worst in my normal usage, everything is great.

So in a word, stable is passing ALL of that and being able to game to your heart's content. If you are crashing in anything, then you have a problem. My previous x58 build ran 4.0ghz for 7 years 365 days a year. I stressed it HARD at the beginning, p95, orthos, intel burn test, all running at the same time for 48+ hours. It passed and I relied on this rig from Jan 2010, to Feb 2017.

For the hell of it, I've been running p95 all day since I first posted, I'm hovering under 90's now, 87-88 on core1, while the others are dangling around 85C It looks like my thermal paste has settled in since I first built this system a few weeks ago.

my ususal strees testing hasb een my 8h/8h/8h test.
8h linpack
8h prime95 + furmark
8h memtest

that way all parts of the system is tested thoroughly

I've never been fan a fan of OCCT i just don't like the slow start and general layout. but thats purely a cosmetic preference



Once i get time I'm going to look into more system wide test AKA CPU+GPU. im just to busy with other projects currently

 

[Kamber]

Cheers, going to aspire to hit this with me 7700k!

 

[Iratus]

I'll give it a go tomorrow.

Mine is cool and low voltage ([email protected]) and almost glitters it's so golden so if it doesn't then I don't think many will.

Wouldn't meet Sven's strict definition of stable though as I only used weak tools (all running at the same time).

And yet I've been doing complex analytics, reencoding my blu rays, playing tons of games and running VMware without the slightest problem. Literally the least hassle I've ever had with a home pc. Go figure.

I did get a good laugh at the passion though, that's a man that takes his stability seriously