Haswell overclocking discussion.

BallaTheFeared

Diamond Member
Nov 15, 2010
8,115
0
71
We seem to be in the phase where people have chips, but nobody is really talking about them nor is anyone releasing any comprehensive all encompassing guides.

I thought perhaps in the meantime we could have an open discussion about what we are doing and what is and isn't working for us as actual users of the product.

First I would like to say that some things I've heard have not worked for me, such as System Agent voltage, Processor Input voltage, bootstrap function (simply fails to post and reverts for me), messing with anything other than what I'll list below as far as what is working.

Well then what is working? Testing methodology, fixed voltage, dynamic clock speed, specified voltage (not offset), and ADIA64.

How to test, and in what order? For me before becoming thermally limited, given other things to do away from the PC ADIA64 seems to be the way to go. Otherwise just gaming in demanding titles like Crysis 3 has produced similar results as well.

The first thing to do is remain free of software based tools, because we'll be relying predominantly on the "Auto" function for the vast majority of our settings we want to stay in bios and let it make adjustments if it sees fit. So then that leaves what voltages to change currently, as far as I know the only thing effecting my overclocking potential at this point is the cpu itself. vcore, uncore, dram voltages are the only settings I manually change, everything else the board handles (except multipliers of course).

Secondly core is priority, the cache is already extremely fast with Haswell, increased cache has a performance benefit, but it is trivial compared to core. Start with core, leave your ram at something low such as 1333 or even 1600, also keep your Uncore at stock x38 assuming you're using a 100 base clock, otherwise whatever multiplier allows you to stay close to 3,800 MHz on the Uncore.

Making adjustments and what does this mean? I don't have all the answers here, but generally when you're overclocking just the core it's pretty straight forward. 124 WHEA errors are going to be vcore for the most part, while straight shut down in power is also most likely not enough vcore. Later on it get's complicated, you can get 124 WHEA errors in my experience when combining all three factors (core, uncore, ram) and not having enough Uncore voltage. I've also seen 101 for not enough voltage (Uncore).

Max multiplier was prevalent with Sandy Bridge, not sure what happened with Ivy I stopped paying attention. However each chip is going to have a max multiplier, with the addition of the bootstrap it might be possible to get more from your chip using this method, I have no idea though because my GD65 or meself seem to be incapable of using the function.

Once you've locked in your vcore setting and core speed you can move on to RAM and/or Uncore, raising uncore voltage here is how to gain stability for the most part, however because your cores are now being fed quicker this may induce some instability with the core, a bump or two in vcore is all you should need here, everything else should come from uncore. Your uncore voltage can't be higher than your vcore voltage last I heard, keep this in mind as for me personally, I ran into this issue when trying to stabilize 4.6/4.6.

Finally order of importance for me and my usage (gaming), Core > Uncore = Ram. Core is king, don't sacrifice core for either ram or uncore unless your specific application benefits for the aforementioned more than others. Also don't be afraid to "settle", silicon is a lottery, sometimes you only win a little when playing the overclockers lottery, but generally everyone wins something.


Step 1: Core focused, only core multiplier and core voltage are changed. Fixed voltage is used to prevent voltage bumps from AVX code.



Step 2: cpuNB /w desired core frequency. You can go for both your desired vram and cpuNB here, but when overclocking it's always best to change only one thing at a time.




In my case I went for cpuNB first, with vram to be tested next with these clocks. However in my testing I was unable to hit 4.8GHz on the cpuNB with the allotted voltage my vcore provided. At this point in my experience you have two options.

Option A) Increase vcore/uncore voltage until the uncore stabilizes.
Option B) Decrease uncore multiplier until it stabilizes.

I chose to go for the lower voltage core clock and drop uncore instead, luckily I was able to verify stability for an hour at these settings with just one notch down on my cpuNB (YMMV).

Step 3: Rolling it all up. In step 3 we take everything we've done in step 1 and 2 and combine it to hopefully achieve our final clocks. This combines our core, uncore, and memory overclock all into one final test to make sure everything ends up jiving with each other at their particular settings.



It's the last step I would personally spend the most time verifying, it has the most factors and is of course the most important. How you go about "stability" testing is completely on you. Each person has their own ideas and opinions on this subject. Since it's a subjective topic everyone is right, do what you do. Personally I've only used AIDA because it seems to be suggested a lot. The reality of my situation is that even though I just spent over 5 hours doing this post with all the testing it's meaningless to me. My testing starts now, when I load up a demanding game and verify that it's working and stable. Which leads me to the next step...


Step 4: Real world usage, now that you've put it through the ringer, use it and enjoy it! Make adjustments as you need and if you find a program that is unstable use it as your new stability test and revert back to Step 1!

Remember silicon is the end all be all factor in your results with Haswell, just like the last few generations (there is some variance with different boards, but in respect to what you have, it's what most others would get with the same setup so don't fret). The fact that there just isn't much you can feasibly do to improve what anyone else with the same chip can get isn't new.

Video & Written Overclocking Guides

Video's from JJ (ASUS) on overclocking.

http://www.youtube.com/watch?v=p2XpXzTAq1c

http://www.youtube.com/watch?v=ub6WnHmiIOs

http://www.youtube.com/watch?v=1DqIG_89Mcs

http://www.youtube.com/watch?v=ub6WnHmiIOs

Video from Linus Tech Tips

http://www.youtube.com/watch?v=0CHs5_TdpXE


Written guide from ASUS covering their "ROG" series

http://rog.asus.com/242142013/maximus-vi/maximus-vi-series-uefi-guide-for-overclocking/

Written guide by Sin0822 on Overclock.net, based on Gigabyte but should translate to other boards as well.

http://www.overclock.net/t/1401976/the-gigabyte-z87-haswell-overclocking-oc-guide

Written guide from overclockers.com "3 step guide"

http://www.overclockers.com/3step-guide-to-overclock-intel-haswell



Programs and Utilities

Wand3r3r's WHEA checker

http://forums.anandtech.com/showthread.php?t=2312638

CPUz 1.64.3

http://www.mediafire.com/?zp28fy8ru48x7r6

RealTemp (TechInferno Edition)

https://mega.co.nz/#!iQBFTRDC

AIDA64

http://www.aida64.com/downloads

Prime95

http://www.mersenne.org/freesoft/

IntelBurnTest

http://www.softpedia.com/get/System/Benchmarks/IntelBurnTest.shtml

Intel® Optimized LINPACK Benchmark

http://software.intel.com/en-us/articles/intel-math-kernel-library-linpack-download
 
Last edited:

JoeRambo

Golden Member
Jun 13, 2013
1,814
2,105
136
Do you have a source where it is claimed that uncore voltage should stay below vcore voltage? I find it a bit hard to believe as it is on completely different voltage plane.
 

BallaTheFeared

Diamond Member
Nov 15, 2010
8,115
0
71
I thought I heard it in one of the video's I watched, either on tek syndicate or newegg, but I think JJ from ASUS was the one who said it though I can't be sure and finding it might be hard.

Perhaps I heard wrong, or the fact that the vcore and uncore change color at difference voltage levels put the thought into my head, but that seems a bit weak. I'll see if I can dig it up if it actually exists.
 

24HZ

Member
May 25, 2013
52
0
0
You know, I got much higher temps on my 4770K with prime95 small FFT than I did with AIDA64. You might want to try that as your temps seem pretty low under load.
 

BallaTheFeared

Diamond Member
Nov 15, 2010
8,115
0
71
Updated OP with what I consider Step 3 and 4, please feel free to add your own insights that's why I made the thread!


You know, I got much higher temps on my 4770K with prime95 small FFT than I did with AIDA64. You might want to try that as your temps seem pretty low under load.

Use what you like, you don't need me to call your chip stable for you to feel your chip is stable :thumbsup:

Temps seem consistent though.

 

WhoBeDaPlaya

Diamond Member
Sep 15, 2000
7,414
401
126
You know, I got much higher temps on my 4770K with prime95 small FFT than I did with AIDA64. You might want to try that as your temps seem pretty low under load.
4770K (HT) vs his 4670K (no HT). Turn off HT on your's and watch temps drop.
 

WhoBeDaPlaya

Diamond Member
Sep 15, 2000
7,414
401
126
No. Struggled with that for a little while, then said forget it.
Happy with the 45 x 101 (~4.55GHz) I was able to achieve.
 

Concillian

Diamond Member
May 26, 2004
3,751
8
81
Updated OP with what I consider Step 3 and 4, please feel free to add your own insights that's why I made the thread!




Use what you like, you don't need me to call your chip stable for you to feel your chip is stable :thumbsup:

Temps seem consistent though.


Minimum 17C... just how cold is your house? Winter isn't coming... it's already there.
 

pm

Elite Member Mobile Devices
Jan 25, 2000
7,419
22
81
Minimum 17C... just how cold is your house? Winter isn't coming... it's already there.

<laughs>


As far as the original, I'll add up my data when I get a chance. I've had some memory compatibility issues that I've been messing around with.
 

Eureka

Diamond Member
Sep 6, 2005
3,822
1
81
Just got my 4770K overclocking... so far it seems stable at 4.4 GHz with a motherboard offset of +.01.

However, I can't figure out what my voltage is doing. CPU-z is reading .9V, and HWInfo64 seems to say 1.2V regardless of my offset being at .01 or .06.

Anyone having this issue?
 

Eureka

Diamond Member
Sep 6, 2005
3,822
1
81
Well, I'm playing around still, but I was able to bump up the adaptive voltage max to 1.25V. Even then, I can't stabilize it at 4.5 GHz, when I ran Intel Burn Test the voltage spiked to 1.33V, and then shut down (no BSOD or freezing) and restarted.

At 4.4 GHz, it's stable even at 1.2V fixed. :/

It's odd that at 1.33V I can't stabilize 4.5/4.6 GHz, although it may also be a motherboard current limit issue (I can't find the current limit setting yet).

I'm going to have to read more on the overclock.net post, but this is good times so far.
 

BallaTheFeared

Diamond Member
Nov 15, 2010
8,115
0
71
Normally for me if it just shuts down you don't have enough vcore :\

What are you using for your uncore and memory at for 4.5GHz?
 

Eureka

Diamond Member
Sep 6, 2005
3,822
1
81
Normally for me if it just shuts down you don't have enough vcore :\

What are you using for your uncore and memory at for 4.5GHz?

I've left those untouched... I'll get to it later when I restart my computer.

The only things I've changed:
Voltage mode to adaptive
Adaptive voltage max to 1.25V
Multiplier to 44
Load Line Calibration to Step 1
 

Eureka

Diamond Member
Sep 6, 2005
3,822
1
81
Uncore is matching my CPU clock... so 4.4 Ghz at the moment. Memory is untouched from stock (10-10-10-27 1600 @ 1.5V). Vin is 1.2V, Vcore is .896V.
 
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/    |