QUOTE(Naraw @ Jun 10 2013, 09:25 PM)
How long to stress test?
I'd give it 3 hours minimum. The longer the better of course.Haswell, Devil's Canyon, Pentium G3258 OC V1, 4790K/4760K/4770K/4670K/G3258 | LGA1150
|
|
 Jun 10 2013, 09:59 PM
Return to original view | Post
#41
|
 
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan   
|
|
Quote|
|
|
|
|
Jun 11 2013, 10:16 AM
Return to original view | Post
#42
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
QUOTE(sam60 @ Jun 11 2013, 09:41 AM) Just to share. so far this is stable for me Nice try. But I can see that you are using an older Prime95 which doesn't really stress the Haswell to the correct level.[attachmentid=3481837] Pls use the latest Prime95 27.9 posted in the first post of this thread, which has AVX library. |
|
|
Jun 11 2013, 10:53 AM
Return to original view | Post
#43
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
QUOTE(chidorizzz @ Jun 11 2013, 10:25 AM) Guys, im stuck at 4.1Ghz.. what is the symptom? Unable to post or stuck at Windows logon screen?maybe my RAM prob? im using Kingston hyper x KHX1866c9D3K2/8GX my mobo is maximus vi extreme, i74770k, noctua NH-D41, Please help.. The most suspected cause is insufficient vCore. Try upping it slightly to 1.1v etc. |
|
|
Jun 13 2013, 10:52 PM
Return to original view | Post
#44
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
ASUS T-Topology 2.0 is working fine here testing 2 vs 4 modules
2 modules -- 6m 17.871s  4 modules -- 6m 16.562s  Same as T-Topology 1.0 on Z77, 4 modules (populated DIMM) is faster than 2 modules, at the same memory settings and voltages.  QUOTE 2nd generation T-Topology 5%+ full load OC margin and 10%+ with one DIMM For extreme memory overclocking under full load, ASUS developed the world’s first 2nd generation T-Topology. It further optimizes layout to minimize coupling noise and signal reflection, increasing DRAM overclocking. Testing shows around 5% more OC margin under full load, and 10% OC margin growth for one-DIMM configurations. Thanks to 2nd generation T-Topology, Maximus VI Extreme sustains RAM speeds over 3000MHz. This post has been edited by owikh84: Jun 13 2013, 10:53 PM |
|
|
Jun 14 2013, 06:38 AM
Return to original view | Post
#45
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
4x 4GB 2800MHz 11-14-14-25-1T is prime95- stable for 3 hours.
 |
|
|
Jun 14 2013, 11:04 AM
Return to original view | Post
#46
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
QUOTE(cstkl1 @ Jun 14 2013, 08:03 AM) is this 1.65v set?? BIOS set is 1.66v, DMM reads 1.67v. Anything lower will result in error in HyperPi unless if the subtimings and command rate are loosen then only can reduce the vdimm etc.cause why is it running more than 1.65v. hmm funny i cant flash to shamino bios anymore. i tested it in the beginning and it sux and when back to 0503. and now wanted to go back but cannot. QUOTE(sam60 @ Jun 14 2013, 10:05 AM) may i know what type of cooler u using? 4.6 ghz with less 80 temp and 1.4 Vcore. cool It is a custom watercooling kit consisted of EK Supreme HF plate #6 CPU waterblock, 2x XSPC RX360 radiator, Swiftech MCP655 pump, EK FC7970 GPU waterblock, EK Multioption X2 Advanced reservoir, XSPC 1/2" ID 3/4" OD white tubing, 12x Scythe GT AP-15 1850rpm 120mm fan + Aerocool Shark 120mm fan, Coolaboratory Liquid Pro thermalpaste etc. |
|
|
|
|
|
Jun 14 2013, 11:54 PM
Return to original view | Post
#47
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
4 X 4GB 2800 11-14-14-25-1T @ 1.63v (BIOS), 1.646v (AI)
 HyperPi:  Prime95 27.9 w/ AVX:  This post has been edited by owikh84: Jun 15 2013, 09:31 AM |
|
|
Jun 15 2013, 09:36 AM
Return to original view | Post
#48
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
QUOTE(cstkl1 @ Jun 15 2013, 02:28 AM) Still dead. Proceeding with removing cpu. Holy cow, so fast! Update still dead. Was told that my cpu might be dead.  So it is the 0017 causing the death or something? Get another CPU or mobo to test? The debug code could be "7B" also. Swap the CMOS battery and see. This post has been edited by owikh84: Jun 15 2013, 09:48 AM |
|
|
Jun 15 2013, 10:28 AM
Return to original view | Post
#49
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
QUOTE(cstkl1 @ Jun 15 2013, 01:32 AM) Bios 0017 awesome. I think you have your CPU killed by excessive VCCIN bro.Mobo dead n stuck n code 78. Was just priming 2250c7. Before that already noticed my g19 was behaving oddly during bootup. Second bios also cannot boot. Clear cmos now with no power n waiting 20mins Shld be ok after this. Odd same timings were ok on 0503. I heard loud_silence and Dumo killed their good chips with 2.0v+ under LN2 and regular air/water. But your situation is worse, cause you are running it on air and torture it with stress tool. If you want to refer to Intel Datasheet for Haswell, the VCCIN is 1.7v default without limit. But i will play anything higher than 1.8v afaik the highest VCCIN I put was 1.75v then slowly playing with auto as recommended by ASUS.  |
|
|
Jun 17 2013, 12:10 AM
Return to original view | Post
#50
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
3000MHz 12-14-14-25-2T @ 1.65v
 RAM clock looks good for 24/7 and priming but CPU clock is nono. This post has been edited by owikh84: Jun 17 2013, 12:14 AM |
|
|
Jun 17 2013, 05:27 PM
Return to original view | Post
#51
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
QUOTE(Naraw @ Jun 17 2013, 05:00 PM) Some noob question. I oc-ed my i5 to 4.5ghz at 1.2v. When I stress test using msi's tool its stable but prime95 not stable. Is my voltage not enough? Currently there is only one reliable stability app for Haswell which is prime95 27.9. Custom blend. |
|
|
Jun 17 2013, 06:03 PM
Return to original view | Post
#52
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
QUOTE(Naraw @ Jun 17 2013, 05:39 PM) I used custom blend with the settings you recommended on first page. It'll be fine for 15mins but longer than that crashes. It could be CPU or RAM setting issues.Try not to touch the memory settings but let it auto then focus on CPU clocking first. Increase your vCore necessarily. |
|
|
Jun 17 2013, 08:06 PM
Return to original view | Post
#53
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
QUOTE(Naraw @ Jun 17 2013, 07:54 PM) I tried again with the same settings for prime95. If more than 15min bsod appears. Ram settings at auto. Try increasing your vCore? 1.3v-1.4v? Just in case if your chip is a lemon |
|
|
|
|
|
Jun 17 2013, 08:43 PM
Return to original view | Post
#54
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
|
|
|
Jun 17 2013, 09:37 PM
Return to original view | Post
#55
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
QUOTE(cstkl1 @ Jun 17 2013, 09:19 PM) Thats amazing I think they are binned.Btw those 3k kits... What are the chipset. Cause hmm corsair qualify them as vengence pro so guessing its not a binning issue. So cant be that rare. 3000C12 kits are Hynix, same as my 2800 11-13-13. But they can't do C10 @ 2800+ which Samsung did. I have a Corsair Dom Platinum 2800 11-14-14 but they can't stable @ 3000. C-N uses Samsung kit of 2666C10. I also have 2666C10 but G.Skill TridentX. A good Samsung should be able to clock 2800C9, but of course at high vdimm. QUOTE(cstkl1 @ Jun 17 2013, 09:20 PM) Btw have u thought of doing like the gamer profile. hmm. I prefer 5.0 running on all cores.5ghz one core etc |
|
|
Jun 17 2013, 09:53 PM
Return to original view | Post
#56
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
QUOTE(Naraw @ Jun 17 2013, 09:43 PM) Just stressed the cpu only using msi's extreme tuning utility and it managed to go pass 30mins at 4.5ghz-1.2v. But when I stressed the cpu only using prime95 it crashes after exactly 15min. I used the max fft setting for prime95. Both stress test ram are default. that means MSI Extreme Tuning Utility is not as stressful as Prime95. So, better don't use it for stability test. |
|
|
Jun 17 2013, 09:55 PM
Return to original view | Post
#57
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
QUOTE(kmarc @ Jun 17 2013, 09:45 PM) Arrrgghhh!!! All my OC for nothing! Let me guess.Booted up my normal windows at 4.2Ghz @ 1.1v. Run prime95 (27.9) and noted vcore jumped to 1.198v and temperature jumped to 81-82'c!!!! Compared to my previous OC (in a fresh windows 7), the weird thing was that the same prime95 ran small FFTs with type-0 FFT. Now in my main windows 7, it is using AVX FFT!!!  Summary Prime95 (27.9) in fresh win 7 (type-0 FFT) - OC 4.2Ghz @ 1.1v, vcore 1.129v in CPUz, realtemp 72'c max (2 hours stable) Prime95 (27.9) in main win 7 (AVX FFT  ) - OC 4.2Ghz @ 1.1v, vcore 1.198v in CPUz, realtemp 81-82'c (in a few seconds)!!!! Wasted my whole day.....  Your main win7 is Service Pack 1 (SP1) fresh win7 is vanilla (none-SP1) SP1 has AVX instruction which vanilla doesn't. |
|
|
Jun 17 2013, 10:07 PM
Return to original view | Post
#58
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
QUOTE(Naraw @ Jun 17 2013, 09:58 PM) Really? I'll try to oc it 4.4 or lower and see whether is it stable. Could it be ram issue? If you didn't touch anything on the memory settings including timing, vdimm, VCCSA, VCCIO then it shouldn't be the RAM issue.Just pump the vCore or downclock. |
|
|
Jun 17 2013, 11:38 PM
Return to original view | Post
#59
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
Managed to lower down the vCore from 1.44v to 1.4v without any magic
 4.6GHz @ 1.4v  3000MHz 12-14-14-25-2T @ 1.65v passed HyperPi  A good start for prime95.  |
|
|
Jun 18 2013, 06:50 AM
Return to original view | Post
#60
|
|
Elite
8,711 posts Joined: Nov 2007 From: Butterworth, PG / Machang, Kelantan
|
ok done priming for 7 hours.
3000MHz 12-14-14-25-1T @ 1.65v  |
| Change to: |  0.0441sec
 0.33
 7 queries
 GZIP Disabled
Time is now: 28th November 2025 - 05:18 PM |
All Rights Reserved © 2002- 2025 Vijandren Ramadass (~unite against racism~)
Powered by Invision Power Board © 2025 IPS, Inc.