Achivement&Record - Overclocking the X99E-ITX/ac

Print Page | Close Window

Overclocking the X99E-ITX/ac

Printed From: ASRock.com
Category: OverClocking(OC) Zone
Forum Name: Achivement&Record
Forum Description: Share your HWBOT or Futuremark score here!
URL: https://forum.asrock.com/forum_posts.asp?TID=188
Printed Date: 22 Nov 2024 at 12:31am
Software Version: Web Wiz Forums 12.04 - http://www.webwizforums.com

Topic: Overclocking the X99E-ITX/ac

Posted By: Jon
Subject: Overclocking the X99E-ITX/ac
Date Posted: 28 May 2015 at 3:08am

I have done a bit of simple overclocking with the X99E-ITX/ac but haven't gotten into it a lot yet. I just wanted to make a thread for anyone else overclocking to share their experiences.

I have the i7-5820K processor with a Seidon 120V all-in-one cooler.

I was able to use the preset configuration to take it to 3.8GHz and 4.0GHz pretty easily. Beyond that, I have gotten up to 4.5GHz by increasing the multiplier to 45 and core voltage to 1.26V.

I have tried a little to get to 4.6 by upping the multiplier and core voltage up to 1.30 but haven't been successful there (yet).

So...has anyone else tried overclocking with this motherboard? Results?

Replies:

Posted By: bog_xdan
Date Posted: 02 Jun 2015 at 5:12pm

Probably you need more voltage vcore 1.35v? . But the cooling might not be enough. I didn't not have this platform.

-------------
Live free or die hard

Posted By: Jon
Date Posted: 06 Jun 2015 at 4:47am

bog_xdan wrote:

Probably you need more voltage vcore 1.35v? . But the cooling might not be enough. I didn't not have this platform.

I think I have some headroom when it comes to cooling, the hottest I have seen so far is 75C. I have been a little hesitant because I haven't been able to find much discussion of "safe" core voltage for Haswell-E CPUs, specifically the 5820K. I think the Intel I have seen 1.30V or maybe 1.32V as the quoted limit from Intel but I'm curious how far beyond that one can go without running into trouble.

Posted By: glnn_23
Date Posted: 19 Jun 2015 at 9:48pm

I'm using this board with an 8 core xeon and found it overclocks very similar to my previous EVGA Micro and R5E with regards to cpu clock.

OCing memory on this board though is nowhere near as good as the R5E . The evga was even better than both. I have run 3dmark benches with my evga with 16 Gb G.Skill 2400C15 @ 3440Mhz but struggle big time with the Asrock.

Hit my highest cache speed of 4600 with the Asrock but much more stable around 4400.

I think the secret to keeping temps down so far is to run with all Cstates enabled.

I found today that when you manually set the cache speed, the CStates automatically disable themselves somehow.

Also I keep Load Line Calibration on lowest setting 5 and run adaptive voltage for the vcore using the 100 strap of course.

Temps have been very good as I'm using EK Supremacy Evo block with the narrower bracket and 2 x HWlabs Nemesis 360GTX rads.

Idle temps at stock for gpu/gpu are very close to water temps if GPUZ, core temp CPUID HWMonitor and Aquaero6 are accurate.

Running the Titanx at 1.27t volts ,1549/2008 and 8 core cpu at just under 1.4v @4.65Ghz and the max temp for the TitanX was showing 28C.

-------------
Caselabs S5, E5-1680v3, Asrock X99E-ITX/ac, TitanX with EK block, EK Supremacy Evo, HWLabs

Posted By: parsec
Date Posted: 22 Jun 2015 at 1:32pm

Interesting about the C States seemingly disabled when the Cache ratio is manually (increased?) set. That would seem to be an automatic behind the scenes change done by the UEFI/BIOS code. I assume you have C States set to Enabled rather than Auto.

I've seen things like that happen with other settings, such as automatically changing the CPU cooler fan speed to Full On when the CPU ratio was increased. That was on a Z97 board, not your board. That change could be overridden by resetting the CPU Fan speed setting and would not change again. I imagine you checked the C State option settings when you noticed C States were not active.

Something else just struck me, how did you know that C States seemed to be disabled? Was that by monitoring the VCore in a monitoring utility? Meaning the VCore did not decrease at idle when C States should be active?

If that is the situation, the monitoring program you are using is displaying VIDs and not the true VCore. That is common with Haswell processors, are you aware of this? Most hardware monitoring programs cannot display the VCore of Haswell processors, since the VCore voltage is created by the internal Haswell VRMs.

HWiNFO64 can display the VCore of Haswell processors on many different boards. I don't have personal experience with X99 boards so don't know if it will show your boards VCore.

Haswell over clockers generally feel that over clocking the Cache speed provides little to no performance increase.

You may know some or all of this, just sayin' Geek

Posted By: glnn_23
Date Posted: 29 Jun 2015 at 7:57pm

Temps immediately became higher. Idle temps went from under 20C to over 30C. Went into bios to see what was different and there it was.

So every time cache is changed CStates change.

-------------
Caselabs S5, E5-1680v3, Asrock X99E-ITX/ac, TitanX with EK block, EK Supremacy Evo, HWLabs