First OC - i7 875K

bryho

Member
Jun 3, 2010
25
0
0


Hi all - first time OCer. Just build system w/ i7 875K, hyper 212+ cooler, gigabyte GA-P55A-UD3 mobo, and some gskill ddr3 1600 ram.

I turned off turbo mode, set bclk to 160 (instead of default 133) and left the cpu multiplier @ 22x for 3.52ghz @ stock voltage of 1.248. idle temps are posted, and they range from 33-37C.

- Is ~24hours of Prime95 the best way to stress test for stability?
- Did I do anything wrong (should i have adjusted the voltage, should i have not adjusted bclk, etc.) ?
- Can I go much further? Most people say 4ghz on air with this cpu/cooler is no problem, but I am mainly OCing for a significant performance boost with minimal effort. I bought this aftermarket cooler but I do not want to be plagued with stability issues down the line just for 1% on some benchmark.

Thanks for any advice !
 
Last edited:

Zap

Elite Member
Oct 13, 1999
22,377
2
81
You always remember your first time.

Mine was with a 386DX25 overclocked to 27MHz with a crystal swap...

Does your motherboard allow you to raise multiplier? I thought the point of the "K" chips was unlocked multiplier? You can probably easily get another couple hundred MHz, but while 4GHz is doable you will probably have to start jacking voltage and will be getting high temps by then. I had a Hyper 212+ on a Core i7 965 at 4GHz with extra voltage and Prime95 would hit 100°C within minutes.

Speaking of Prime95, my opinion is that if it can run 1/2 hour, then it is good. The theory is that for utmost stability your rig should be able to run it indefinitely without errors. However, who's going to build a system just to run Prime95 forever unless you were actually a mathematician searching for the next prime number? So, some people will say that 1/2 hour is not enough because it could error after 1/2 hour. To that, I say "sure it can." However, even if it is stable for 24 hours, who's to say that it won't error out in the 25th hour? See, it is all kind of arbitrary. Why I choose the 1/2 hour is that usually temperatures will have peaked by then, so not only are you testing for stability, you are also testing for stability under duress of high temperatures.
 

lopri

Elite Member
Jul 27, 2002
13,212
597
126
You didn't do anything wrong and the overclock looks good so far. Congrats. The chip can surely be clocked higher and if you want to take it further you can start by reading guids like this.

As for Prime, I largely agree with Zap. If a chip passes Prime for 1/2 hour ~ 1 hour, then it's very likely the chip is good at that frequency and it will keep running. Failure after that usually comes from changes in ambient temperature. Although to be sure, it is also possible that the motherboard's power delivery system or a power supply give in after a certain period. Various interconnects (QPI, HTT, etc.) will also be affected if power delivery fluctuates and temperature changes.

My favorite routine for final stability test of a given overclock goes like this:

1. Full 2 loops of Memtest
2. 50 loops of Linpack using all available memory
3. 4 hours of Prime LargeFFT
4. Some sort of tests that stress system buses, namely PCIe/HTT/FSB/QPI. PCMark/3DMark Vantage and Crysis do the job.

If you get into memory overclocking somehow, which usually does not provide practical benefit but can be fun and rewarding nonetheless, you'll proably want to run Super Pi and HCI memtest.
 

PCTC2

Diamond Member
Feb 18, 2007
3,892
33
91
I'd say, since you have the i7-875K, use it for what its main selling point is... unlocked multiplier. Leave the BCLK where it is and start stepping up the multiplier. Once you get to an unstable frequency, back it down one multi, and then start bumping the BCLK 1 or 2 MHz at a time until it's unstable again. Then back it down a little and then you've got a really nice OC.
 

Zap

Elite Member
Oct 13, 1999
22,377
2
81
My favorite routine for final stability test of a given overclock goes like this:

1. Full 2 loops of Memtest
2. 50 loops of Linpack using all available memory
3. 4 hours of Prime LargeFFT
4. Some sort of tests that stress system buses, namely PCIe/HTT/FSB/QPI. PCMark/3DMark Vantage and Crysis do the job.

If you get into memory overclocking somehow, which usually does not provide practical benefit but can be fun and rewarding nonetheless, you'll proably want to run Super Pi and HCI memtest.

LOLOLOL on memory overclocking.

Your first two stability tests both stress memory, which is something I've been leaning towards lately. Seems to me that memory... just isn't so robust these days. I've had more problems getting memory to even run at rated specs than I've had with clocking CPUs high.
 

jvroig

Platinum Member
Nov 4, 2009
2,394
1
81
If a chip passes Prime for 1/2 hour ~ 1 hour, then it's very likely the chip is good at that frequency and it will keep running. Failure after that usually comes from changes in ambient temperature. Although to be sure, it is also possible that the motherboard's power delivery system or a power supply give in after a certain period
Absolutely. If the Intel chips do not behave too differently from AMD ones, then sometimes reboots happen after 3 hours of Prime95, which can be cured with a bump in vcore.

Still, I would agree with Zap. For most normal computers, it's likely that even just an hour of Prime95 is already an unreasonably heavy load, as the intensity and duration of the loading is something that most likely won't happen in real world usage anyway, especially for quad cores. Whether it's a reboot or an error detected by the stress testing suite, it's something that most likely will never happen in real usage anyway if it takes several hours of a CPU load what you would not experience in real world usage.

Folders and heavy video encoders are probably a stark exception.
 

bryho

Member
Jun 3, 2010
25
0
0
Thanks for the followups.

Small issue that has me a bit confused...I went into the tweak section of BIOS and went to increase the CPU multiplier past 22x (currently 22x160), but it will not let me go past 22x. Shouldn't I be able to set the multiplier to anything I want? Is there a setting I need to tweak to tell it that I have an unlocked CPU?
 

bryho

Member
Jun 3, 2010
25
0
0
Thanks for the followups.

Small issue that has me a bit confused...I went into the tweak section of BIOS and went to increase the CPU multiplier past 22x (currently 22x160), but it will not let me go past 22x. Shouldn't I be able to set the multiplier to anything I want? Is there a setting I need to tweak to tell it that I have an unlocked CPU?

Anyone able to help with this? Quite frustrating =(
 

jvroig

Platinum Member
Nov 4, 2009
2,394
1
81
Not an Intel owner, but seems to me from what you have said that you have reached the max multiplier, so now you need to raise the FSB / base clock / bus speed, which is currently only at "160", in order to get past 3.5GHz
 

MeTechE

Junior Member
Apr 3, 2010
20
0
0
I have two 875k's (selling one here) and I'm still waiting for my motherboard to arrive so I can't speak from experience yet. It looks like you guys found the Legit Reviews article where they got it to 4.49Ghz. Someone on another forum got theirs to 4.2GHz+ (Linx stable) with 25x 165.5. Legit Reviews used a TRUE for cooling, which may have made the difference.



I hope to join you in overclocking these chips very soon .
 
Last edited:

David K666

Junior Member
Jun 23, 2010
2
0
0
Hi there this is David from another forum and was curious.... when you use a Gigabyte P55 or H55 motherboard with the i7 875K unlocked cpu does it give you access to any " Turbo Boost ratios" or anything else that would be essential for raising solely the multiplier to achieve a higher OC???

Or could I still do it (say from 22x to 25X) and not have to worry about that????

Cause on a cheaper motherboard I wouldn't expect the cooling (if any) to hold up anyway....:\

and All I would want is 3.33 ghz for daily use without messing with the bclk...

Thanks

David
 
Sep 22, 2009
65
0
0
all gigabyte owners need F8 or newer betas for higer than 22 multi. but that said seems those chips are voltage hogs over 4gz compared to the 860's and stanrd 870's. but up to 4ghz they seem to run fine. I'd say bump the core voltage up to 1.3v and take it as high as she'll go
 

MeTechE

Junior Member
Apr 3, 2010
20
0
0
Hi there this is David from another forum and was curious.... when you use a Gigabyte P55 or H55 motherboard with the i7 875K unlocked cpu does it give you access to any " Turbo Boost ratios" or anything else that would be essential for raising solely the multiplier to achieve a higher OC???

Or could I still do it (say from 22x to 25X) and not have to worry about that????

Cause on a cheaper motherboard I wouldn't expect the cooling (if any) to hold up anyway....:\

and All I would want is 3.33 ghz for daily use without messing with the bclk...

Thanks

David

I will be able to confirm this later but I'll take a stab at it now anyway. The 875K's default multi is 22, if you raise that to 25 (which shouldn't require any more voltage) you will have your 3.33Ghz. AFAIK single threaded turbo boost on this cpu would add 5x multiplier to whatever multi you have set (making the 3.33Ghz 3.9Ghz in single threaded tasks). This should work because turbo keeps the TDP the same by shutting down unused cores (thus requiring no change in voltage). I'll give this a try tomorrow when my motherboard arrives.


I'm selling another i7 875K here, if anyone wants to get in on the overclocking action .
 

David K666

Junior Member
Jun 23, 2010
2
0
0
thanks

I appreciate it...

the ONLY reason I would put the 875K on such a cheap platform as the H55M S2H is to take advantage of the unlocked multi and not have to fool with the bclk and other sources of heat...

upping the multi to 25 and turning 2.93ghz into 3.33 ghz with the full 8 threads should almost be a given...
 

sxr7171

Diamond Member
Jun 21, 2002
5,079
40
91
You always remember your first time.

Mine was with a 386DX25 overclocked to 27MHz with a crystal swap...

Does your motherboard allow you to raise multiplier? I thought the point of the "K" chips was unlocked multiplier? You can probably easily get another couple hundred MHz, but while 4GHz is doable you will probably have to start jacking voltage and will be getting high temps by then. I had a Hyper 212+ on a Core i7 965 at 4GHz with extra voltage and Prime95 would hit 100°C within minutes.

Speaking of Prime95, my opinion is that if it can run 1/2 hour, then it is good. The theory is that for utmost stability your rig should be able to run it indefinitely without errors. However, who's going to build a system just to run Prime95 forever unless you were actually a mathematician searching for the next prime number? So, some people will say that 1/2 hour is not enough because it could error after 1/2 hour. To that, I say "sure it can." However, even if it is stable for 24 hours, who's to say that it won't error out in the 25th hour? See, it is all kind of arbitrary. Why I choose the 1/2 hour is that usually temperatures will have peaked by then, so not only are you testing for stability, you are also testing for stability under duress of high temperatures.

Agree with your Prime 95 philosophy.
 

lopri

Elite Member
Jul 27, 2002
13,212
597
126
Is there a consensus wrt 875K's? I know there are a couple trivial features missing for desktop usage, but otherwise not distinguishable from regular 870 except unlocked multiplier? (which I'd take as a good thing)
 
Sep 22, 2009
65
0
0
I guess they are the last lynnfield chips intel had made and due to 1156 hitting EOL soon, they simply altered the chips by unlocking the multi thus making them cheap, whereas when the original QX9XXX xhips came out they were a completely new architechture hence the more expensive cost and intel saw they would sell that way so they kept them high. but the original unlocked chips QX6XXX chips were not as expensive but still wuite high. just shows intel can simply unlock ships and sell them as low as they want, a big FU to AMD if you ask me, but also to us because of the low binning on the chips
 

Nixxx

Junior Member
Sep 13, 2010
1
0
0
I'm using i7 875K for 2 weeks on GA-P55A-UD5. Due to my memory frequency (2000 MHZ) by BCLK is 167 Mhz, with multiplier set to 20 it should reach 3340 MHZ. But rising a BCLK the Turbo Boost is always on and multiplier is constant and set to 22 what gives 3674 Mhz. That is strange for me, did someone experiensed such a behavior? When it is about temperature in Idle the core temps are around 45. When I start a software like linx or intelburn test the core temps can reach 99C after few minutes. It also happens when I lower frequency to 3300 MHZ. During the game or normal work its normal around 60 C. Does it have any sens to use such programs?
 

StrangerGuy

Diamond Member
May 9, 2004
8,443
124
106
Is there a consensus wrt 875K's? I know there are a couple trivial features missing for desktop usage, but otherwise not distinguishable from regular 870 except unlocked multiplier? (which I'd take as a good thing)

875K OCing should be better since you can choose not to mess with the BCLK which affects the on-die PCIe controller.
 

betasub

Platinum Member
Mar 22, 2006
2,677
0
0
Yeah, forget about running your memory at such a high speed (even if it rated so) when you can use multilier overclocking and Turbo to get just as good results with BLK at default (133MHz) or only slightly higher.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |