• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Overclocking 4820k

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Lightz39

New Member
Joined
Apr 7, 2015
Hello everyone, I've been reading around how to best overclock my system. First my specs:

Motherboard: Asus P9X79 LE
Processor: Intel Core i7 4820k @ 3.7ghz
Memory: 16GB (4x4) 1866 Corsair Dominator Platinum
Video Card: Nvidia Geforce GTX 780ti Two way SLI
Cooler: Corsair liquid cooler.
Powersupply: Corsair AX1200i

I've had success using the auto overclocker that comes with my mobo but it sets the voltage to over 1.4. And changes the blck up from 100. I'm looking to hit a stable mark at 4.5 - 4.7. Does anyone have any suggestions on what to work with to push my cpu to its limits?

I'm pretty new to overclocking. I'll want to set a voltage offset of like +0.75 or whatever to hit 1.4 or less? Sync all cores to 46 or 47? What are some settings I should disable and enable to help with ocing?

Thanks for any input.
 
Welcome to the OCF,

Best is to use manual OC. Set CPU ratio to x42 , CPU voltage to 1.25V and check if it's working fine. No offsets etc, at this stage you test what your CPU can do.
When you make it work at 4.2GHz then you set x43 ratio and test the same , one more up etc till CPU lose stability. When you see it's not stable anymore then you raise voltage, let's say to 1.3V and check higher ratios.
I wouldn't go above 1.35V as temps are going up much faster above some point and I doubt you will overclock it much higher.
Auto OC is almost always setting too high voltage and generally are problems with stability so it's not recommended to use it.

To keep stability you have to raise voltage, lower temps or both. Worse part is that raising voltage you raise temps and every higher memory clock has lower tolerance on high temps.

Let's say it will run like this:

4.2GHz 1.20V
4.4GHz 1.25V
4.6GHz 1.35V
4.8GHz 1.45V

but to keep it stable you will need something like:

4.2GHz up to 90*C
4.4GHz up to 90*C
4.6GHz up to 80*C
4.8GHz up to 70*C
5.0GHz up to 50*C ( under load )

That's just theoretical example and I hope you know what I mean.
Air/water cooling can barely keep these CPUs stable @4.7GHz.
 
Last edited:
I actually don't have the corsair water cooler anymore. I copied and pasted that from an old spec list. I currently have http://www.xs-pc.com/watercooling-kits/raystorm-twin-d5-rx240-v3-watercooling-kit

So far my findings: Overclock set to XMP, core ratio to 46 and voltage offset at 0.090 I have what appears to be a stable machine. I have only done a few quick tests. Nothing long yet. My max temps at full load are sitting below 50c and my voltage at 1.3-1.35. I can't for the life of me get it to run at 4.7 stable though. Does it really require that much more voltage to move up 0.1?
 
Good choice on the cooling :thup:

Set the voltage manually to 1.4V instead of using the offset and then attempt 4.7.
What are you using for stress testing?
 
Thanks! I seem to have lost stability since switching to offset. I was hoping to save a little life on my cpu doing that but 1.4 was much more stable. Even at 1.4 my system will hang as soon as I start running prime.
 
That means you aren't actually stable at 46x. Bump it down to 45x.
 
I think I said that in a confusing manner.
So at 46 and a manual voltage of 1.4 I'm stable. (As stable as a 20 min prime says for now) If I set my voltage to an offset it loses stability. I tried 47 at the manual 1.4 as well and that also isn't stable. There isn't any other way to push it up to 47 without going over the 1.4 voltage is there? I'll try a 8 hour prime test on 46 and manual 1.4 and see how that goes.
 
Ok thanks for the support. I'll post back after doing my long stress test. Then it'll be on to ocing my graphics cards.
 
At 46 and a manual vcore of 1.4 I successfully passed a 2ish hour intel burn test and an overnight prime95 test.
 
Yeah that's the plan. I'm happy with that. I would have liked to hit a little higher but I'm not risking frying anything.
 
4820k 4.7Ghz @ 1.293Vcore 100% p95 load. When idle it drops slightly to 1.26Vcore. Max temps 50 degrees under 100% load idle is little above ambient room temp when settled. Asus P9X79 MB, 4820k on a h110i gt. Will post pics and bios settings if anyone is interested? Have been running this overclock for around 5 months now after a 24 hour p95 burn in. It is rock solid stable and is yet to let me down stability wise.

I modified the cooling on the board by strapping a fan to it above the vrms and caps. As a result cpu power phase being set to extreme never brings me above 32 degrees board wise even when my two GPUs are going at it 100% (with it being an exhaust area they do impact it slightly tho it is cooler just doing p95 run more 29 degrees). This setting ''cpu power phase'' basically means the capacitors that feed power into the cpu don't take breaks and consistently hold power at the correct frequencies even when its not needed so much based on load. It has allowed me massively reduce core voltage on this cpu. Between this and a few other settings ive managed to really bring down my voltage and still get a nice clock speed (mainly this tho). Not many people really talking about this kind of thing on 4820k clocking forums at least? Wondered what your guys thoughts were and whether or not this is old news to you? Maybe your all already doing it? But after my experience with this chip these voltages on here seem very high by comparison to mine? Or maybe ive just been very lucky?
 
Hello yesterday i was overclocking this CPU @4.5Ghz i was able to login in windows and passing 10 sessions of Intel Burn Test in HIGH mode @1.1995vCore, is this normal at this low voltage to get passed the test? In game i was not stable i have to do sure more tests in this days, hottest core was 65 degrees

any consideration ?

[email protected] Hottest Core 72°
unnamed (1).jpg
[email protected] Hottest core 65°
unnamed.jpg

thank you so much for the consideration, i'm from Rome Italy sorry for bad eng ;)
 
Last edited:
72c is still well below any temp level concerns. You should bee good all the way of to 90c on the hottest core. So push that vcore some more if necessary to get her stable.
 
72c is still well below any temp level concerns. You should bee good all the way of to 90c on the hottest core. So push that vcore some more if necessary to get her stable.

Thanks for the answer, I just feel that this CPU is quiet lucky with voltages, or perhaps the Asus Rampage 4 extreme is easy to OC :)

The i7-4820k CPU is soldered to the IHS or there is thermal **** paste between CORE and IHS?

I suppose it should be soldered..


 
Core i7 4820K on MSI X79A-GD45 Plus motherboard

Hi everyone,

I'm currently trying to overclock my CPU on MSI X79 motherboard with auto settings, but for some reasons I'm able to get about 4,2 GHz (mainly changing base clock from 100 to 105MHz). Could anyone recommend some manual tweaks to the system to get something around 4,5 GHz? Just also want to mention that current cooler is able to handle a lot, but I have got CM Liquid Pro as well I'm going to use very soon. Will be very grateful for quick responses. Below is my system conf.

Motherboard: MSI X79A-GD45 Plus
Processor: Intel Core i7 4820k @ 3.7ghz
Memory: 16GB (4x4) 2400 Kingston HyperX Beast
Video Card: GB Geforce GTX 1060 6GB
Cooler: GamerStorm Lucifer.
Powersupply: Corsair RX650

Thanks
 
Overclocking i7-4820k on Gigabyte X79-UP4


My experience has been that the 4820k will over clock to 4.3ghz, just by turning up the core ratio. Though 4.3ghz isnt really stable, it will crash every 3rd or so day (i found a little bit more vcore got it stable). But 4.2ghz will be rock solid, with out vcore tweaks.

There is a few ways to get into the 4.4ghz - 4.6ghz range. I was able to get 4.5ghz stable by turning up the core ratio to 45, increasing vcore to about 1.37 - 1.38 and increasing the Line Load Calibration offset (this got it stable, when it was under load). The highest temp i recorded was 58c, I am running a Corsair H110i.
I found just increasing the core ratio wasnt stable when i tried for 4.6ghz. So to get to 4.6ghz, I increased the base clock, wound back the core ratio to around 44-45, increased vcore to 1.4, increased the vcore response rate and load line calibration and turned off the power saving states. Since I had increased the base clock, I gave my ram a few tweaks and over clocked it from 1600mhz to 1930mhz. The system was stable and temps were around the 62c mark.

I tried for 4.7ghz, but just couldnt get the system stable. Though i was able to run the first 2 cores at 4.7ghz and the last 2 at 4.6ghz and have a stable system.

From what i have read most 4820k's will land in the 4.4ghz - 4.6ghz range, few will land in the 4.7ghz+ range.

Cause i run my system 24/7 I wound the OC back. I run the first core at 4.6ghz then the rest at 4.5ghz, vcore is at 1.38 and temps are around the 50c under load. I also have my ram oc'ed at 1866mhz (from 1600mhz)

If you are having stability issues a few things to try are: increasing the Line Load Calibration (this will stop voltage sag across the mother board when under load), dropping the last core down by 0.1ghz and more vcore.
I think disabling the power saving states will also help with stability and so will increasing the responsiveness of the CPU power system, but i am not 100% sure. I have had positive but mixed results, when playing with those settings.
Also avoid any dynamic vcore settings, I found this was responsible for instability.


"Eleanor" - Built April 2014 - GPU upgrade in 2015 (from R9 280X - 4gb)
MotherBoard: Gigabyte X79-UP4
CPU: i7 4820k @ 4.5ghz
RAM: 32GB (8x4) 1600 G.Skill Ripjaws @ 1866mhz
GPU: MSI R9 390X - 8GB @ Core: 1130mhz Memory: 1600mhz
Cooler: Corsair H110i
Case: Cooler Master Storm Scout 2
PSU: EVGA SuperNova 650 Watt
Storage: 2x2tb WD Black HDD, 1x120gb Kingston SSD, 1x400gb Samsung SSD
OS: Windows 10 Pro 64-bit
3D Mark Time Spy:
 
Last edited:
Back