Overclock.net › Forums › Intel › Intel CPUs › Haswell Overclocking Guide [With Statistics]
New Posts  All Forums:Forum Nav:

Haswell Overclocking Guide [With Statistics] - Page 126

post #1251 of 19540
Just wanted to let you guys know...

ANY MSI Z87 GD65 OWNERS:

the newest bios, version 1.4 was just released.
it seems to add a lot more fine tuning with voltages.
there may have been more bios' released for other boards but i didn't check yet.

thumb.gif
post #1252 of 19540
new bios is lookin good with allowing the VCCIN drops

post #1253 of 19540
Quote:
Originally Posted by Takinato View Post

Well I just realized that my case is only able to fit smaller ones like kraken x40 and h80i. The h100i fits with modification to case which I'm willing to do but not sure of the other 240mm radiators.

i did what i had to do. i took a case with no water cooling features....and water cooled it smile.gif

define r4 is a sweet case if it goes on sale.

http://www.youtube.com/watch?v=vNuxXEb-hgA
post #1254 of 19540
Hey dudes, I just wanted to say that this thread has been very useful in my own efforts to overclock my 4770K. Haswell is a different beast unlike the 2500K I had before. It's definitely trickier to o/c. This thread help shed some light on all the different settings. Thanks Darkwizzie!

Here is what I have so far with my 4770K.

CPU: 4770K delidded
Core Mult: 47
Uncore: 42
Vcore: 1.34
Uncore volt: 1.14
Cooler: Seidon 240m
Ram speed: 1600mhz
Batch #: L310B492
Tested with IXTU plus winamp on repeat thumb.gif



BTW... here is my 2 cents about stability testing programs. IMHO the latest Linpack is the toughest stability test to pass.

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

If you can pass Linpack at the highest setting you can pretty much pass any other stability test. It generates the most amount of heat. The latest Prime95 being second.

http://www.mersenneforum.org/showpost.php?p=350674&postcount=238

This Prime95 version uses the FMA3 instruction set. It recognizes Haswell. It also generates a ton of heat. Then you have Aida64 and IXTU. I have not tried OCCT on Haswell yet. Anyways thanks again to Darkwizzie and the other posters for sharing useful info to help fellow overclockers.

Cheers! cheers.gif
Edited by killaho - 8/27/13 at 11:25pm
post #1255 of 19540
Quote:
Originally Posted by Big Texas View Post

Just wanted to let you guys know...

ANY MSI Z87 GD65 OWNERS:

the newest bios, version 1.4 was just released.
it seems to add a lot more fine tuning with voltages.
there may have been more bios' released for other boards but i didn't check yet.

thumb.gif

Assuming it's also an update for the g45, heck yeah!
post #1256 of 19540
gave up Prime95. something is fishy about it. testing with XTU and H.264.

XTU ran for 13hrs until I stopped it. just the CPU stress test though. didn't do the memory stress test. guess that is not needed if i can pass H.264.

started at 10PM last night and ended at 11am today. 13hrs and 02min to be exact.



now running H.264. been running for 4hrs and 15min without a hiccup.

funny thing is, H.264 also heats up the CPU as much as XTU. frown.gif the ambient is pretty high, around 30C and the fans are not running at full speed. but still, that's a bit on the high side, eh?
post #1257 of 19540
I wish I understood more of the "why" of Haswell.

Why does it pass one day with a given set of settings only to fail the next with the same settings? (I think that one may be temp related)

Why does it always seem to fail in pass 9 or 10 or IBT? Why does it always seem to fail in minute 10 or 11 of a Handbrake encode? (maybe also temp related?)

Why does x264 encoding take so much more Vcore for stability than IBT (in my case +0.025V)? What is x264 stressing differently than IBT, and is there a voltage other than Vcore that would fix it? (maybe it is memory controller related?)
post #1258 of 19540
Quote:
Originally Posted by Darkwizzie View Post

9c is a red herring. businesssmiley.png

Anyways, maybe I can start building a small Bsod code list. 101, Vccin, 124 vcore, 9c ???
I second this.

101 - more vccin is needed, can be caused by raising vcore, SA/analog/digital, ring, without raising vccin. Simply not enough input voltage for cpu.
124 - more vcore is needed, but also too much vcore is applied. Can also mean bad frequency/uncore relation. Adjusting ring voltage without touching other settings can help.
9c, 0A, 01A, 0F7, etc - seems memory related (SA/analog/digital), but can also be resolved by aplying more vcore.
post #1259 of 19540
Quote:
Originally Posted by Forceman View Post

I wish I understood more of the "why" of Haswell.

Why does it pass one day with a given set of settings only to fail the next with the same settings? (I think that one may be temp related)

Why does it always seem to fail in pass 9 or 10 or IBT? Why does it always seem to fail in minute 10 or 11 of a Handbrake encode? (maybe also temp related?)

Why does x264 encoding take so much more Vcore for stability than IBT (in my case +0.025V)? What is x264 stressing differently than IBT, and is there a voltage other than Vcore that would fix it? (maybe it is memory controller related?)

Yeah lol

One day it"ll pass the stress, next day it bsod's...puts me and I'm sure anyone else who gets this in a serious state of confusion.
post #1260 of 19540
Quote:
Originally Posted by byardz View Post

Yeah lol

One day it"ll pass the stress, next day it bsod's...puts me and I'm sure anyone else who gets this in a serious state of confusion.

Yup, exactly what happened to me. Heck yeah 8 hours stable on P95 at 4.4GHz,. *Let's system cool down and then plays BF3 on and off for a good 3-4 hours, BSOD? ***?!?!?

Then, I said screw it and pushed passed to 4.5GHz, so far no BSODs! Oh Haswell.....
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › Haswell Overclocking Guide [With Statistics]