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

I would like to get a basic OC on my I5-4690K

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
For now you don't need to do much. Just raise the voltage and put a proper multiplier and control some of the OC with BCLK, but don't go too high on the BCLK or else it will take like 200 watts extra on your load.
 
On that 1 hour test i did i noticed the back side of my case by the VRM was crazy warm
I have a small fan to zip tie to the VRM heat sync this weekend
if my goal is higher clocks why should i use BCLK over 103 when 103 would push it over another 100 mhz
since BLCK changes the clocks on everything it seems kinda odd to push that alot

I think my issue is my VRM is dropping volts cause it is getting hot
 

Attachments

  • NF-A4x10 PWM.jpg
    NF-A4x10 PWM.jpg
    1,010.9 KB · Views: 623
Leave bclk alone. No reason to touch that.

The vrm should be fine with that cpu really...but cooler never hurts. 70C vrm temp isnt bad co sider most are rated to over 100c.

That board shouldnt have an issue with that cpu.



For now you don't need to do much. Just raise the voltage and put a proper multiplier and control some of the OC with BCLK, but don't go too high on the BCLK or else it will take like 200 watts extra on your load.
So.... how will touching the bclk use 200 more watts???? Why would he touch bclk??? Curious advice here...
 
Last edited:
His LLC should have options like normal regular high ultra high and extreme. Yours should have these options ?

Just use offset and OC, let her put the voltage for you. Thanks
 
Last edited by a moderator:
Can you please share why raising the BCLK will 'use more than 200W'?

Why would this user want to adjust BCLK instead of the CPU multipler?


This advice is....questionable here on a number of fronts bud...
 
after a 30min test with 4.4Ghz at about 1.28V and 1.15v x39 cache it did not crash, i guess the cache is probably stable then
gonna give 1.295 @ 4.5Ghz a try

I'm not really sure where the OP is currently at right now with his overclock??

But it appears that his 4690K sample needs ~1.28v/1.30v volts for 4.4GHz... I'd think the chip would also do 4.3GHz with ~1.22v/1.25v vcore and 4.2GHz with ~1.15/1.20v vcore.

If the OP wants to push upwards towards a 4.5GHz overclock then I'd test or look at trying out in BIOS using 45x core settings @ ~1.35v vcore and 39x cache settings @ ~1.15v with CPU Input voltage @ ~1.8v LLC enabled.

The chip might even need a little more vcore if the sample is approaching/climbing a voltage wall?? Also If you are stress testing with Prime95 (AVX enabled) at ~1.35v(+) then you might end up being temp limited and run out of cooling? I'd try doing initial stress tests with something that runs a little less hot like AIDA64 or ASUS RealBench.

The OP could also check/test some of the following BIOS settings from the Z97 Killer User's Manual:

Page 97 of the manual... FIVR (fully integrated voltage regulator) Configuration:

FIVR Switch Frequency Signature: Select whether to boost or lower the FIVR Switch Frequency.

FIVR Switch Frequency Offset: Configure the percentage of frequency boost or deduction.

CPU Integrated VR Faults: Disable FIVR Faults to raise the threshold to trigger CPU over current protection and over voltage protection for better overclocking capabilities

CPU Integrated VR Efficiency Mode: Enable FIVR Efficiency Management for power saving. Disable for better performance and overclocking capabilities.

Often default/auto motherboard BIOS settings work fine but it wouldn't hurt to tweak/try out some of these to see if they might help a little with your overclock.

Usually the biggest factor in good overclock comes down to the individual quality (margin) of the silicon within a particular chip sample.
 
These haswell/devil canyons clock pretty good . With a after market hsf 4.4-4.6 should be doable .

First I would clear cmos , then set XMP for your ram . Then I would try 45x for CPU multi and 40x for Cache . Vcore I would try @ 1.3 and work down from there . 1.3 will prob be to high but it should @ least let you boot .
From what I have seen the DC chips can do 4.5ghz any were from 1.15v-1.3 .
Post back if you need more help also check that guide Earthdog posted its very good .

Shouldn't he just try offset so he doesn't have to worry about setting vcore. Lets get your CPU to 4.4Ghz stable then take it from there. Don't just jump the gun soon take your time in the OC adventure. Im just saying its a harder ball game if you want to control your voltages, Just put offset in BIOS and let her take care of it.
 
If I were you, I would try setting to OFFSET so you don't have to worry about vcore and what to set. Do your OC with multiplier and little with the BCLK. What you trying to shoot for ? 4.4Ghz would be nice.
 
You posting is so damn confusing....


Why do you keep insisting on using BCLK? Why does using BCLK add "200W"? Can you follow a thread and reply, please??? This is getting a bit frustrating, the scatterbrained posting.
 
The VRM will throttle the Processor clock speed or shutdown the PC when usually the VRM is over 100c+. You won't see the VRM lower the volts when it is 100c+.

assuming the board has the proper protection yes, since the system just crashes even on known good configs i slapped a fan on it
beside given this i am not going to assume it actually has OTP

Either way if the VRM is the problem a fan will normally solve it

4.3GGhz failed at about 45min with 1.22v
 

Attachments

  • NF-A4X10 PWM Installed.jpg
    NF-A4X10 PWM Installed.jpg
    1.1 MB · Views: 546
4.3GGhz failed at about 45min with 1.22v



45 minutes long is not too bad. You are probably getting close... Try again with all cores at 43x and bump vcore up to ~1.25v and cache at 39x with cache voltage at ~1.15v. Also CPU input voltage at 1.8v LLC enabled.
 
CPU input at 1.76 (sensors reads 1.81); LLC is off; vcore is now at 1.225; cache is at 35x
every time you change the multiplier, the bios turns LLC off, but it enables LLC for stock, so i think it is mislabeled binary logic, not sure where to probe it with a multi meter to verify that
** Edit: Multi-meter shows 1.81v at the caps on the left side of the cpu
 
Last edited:
45 minutes long is not too bad. You are probably getting close... Try again with all cores at 43x and bump vcore up to ~1.25v and cache at 39x with cache voltage at ~1.15v. Also CPU input voltage at 1.8v LLC enabled.
4.3 @1.225 survived around 1hr 20 min, assuming stable enough
4.3 @1.225 w/ x39 cache at 1.15 failed at around 20 min
4.3 @1.225 w/ x39 cache at 1.155 failed at around 1hr 5 min
4.3 @1.225 w/ x39 cache at 1.16 TODO

Does LLC affect input or core voltage?
 
43 Core;39 Cache; 1.225 vcore; 1.16 vcache; 100.0 Blck seems stable w/ 80min test (101 Bclk crashed at 12 min, expected it but had to try)
*LLC is off
i am using linux and as far as i know there is no CPU-Z equivalent, the closest i have is lm-sensors
The good thing is i just figured out what input (in6) is vcore :)
the bad news is the sensor data only goes to 2 decimal places and IIRC CPU-Z shows 4 or 5
and to keep the good news going i think i found the data that is VRM temp "PECI Agent 0:"

I would guess if my theory of LLC enabled/disabled being inverted if i turn it on i probably will not even get into the OS

Do not feel like swapping HDDs around and waiting on windows to boot of that array

Edit: LLC enabled does boot and did not insta-fail in a stress test
 
Last edited:
CPU Load Line Calibration only has enable/disable options
I think i have had that setting randomly change when changing profiles or rather not change when it should

all the options aside from vcore override voltage and the all core multiplier are based on the included OC profiles that come with the bios (i have no faith in those profiles what so ever, i know the stock 4.4 profile is unstable)

At this point I suggest doing this. Set your vcore to OFFSET. Then do your overclock but don't take the BCLK past 120 and work with multiplier and shoot for 4.4Ghz first see how the offset goes. let us know.
 
Please explain what that (offset and raising bclk) intends to accomplish, organik...

Youve suggested this in multiple threads with no explanation of why this can help (and we asked multiple times to explain it). Knowing how offset works, this wonr help, and there isnt a point to touch bclk....
 
Last edited:
Back