My i7 6700k can't go over 4.5 Ghz?

[jbake47]

I can't get over 4.5 ghz with 1.35v on my CPU. It's not delidded or anything but this seems pretty bad. Did I just lose the silicon lottery or is there something I'm doing wrong?

GPU: MSI RX 480 8gb

RAM: 16GB Corsair Vengeance LPX @3000MHz and 1.35V

Cooler: Noctua-NHD15

Motherboard: Gigabyte Z170X-Designare

 

[Makentox]

There can be many reasons, at least tell us what are u doing to overclock. At this stage all u have to do is rise multiplier and manual voltage, dont adjust anything else yet.

 

[jbake47]

There can be many reasons, at least tell us what are u doing to overclock. At this stage all u have to do is rise multiplier and manual voltage, dont adjust anything else yet.

At this point I am doing nothing but touching the Multiplier. Everything else is left at default/Auto.

 

[burtman88]

Anyways its possible its increasing you ram speed at the same time? could be another raisin?

 

[burnhamjs]

I can't get over 4.5 ghz with 1.35v on my CPU. It's not delidded or anything but this seems pretty bad. Did I just lose the silicon lottery or is there something I'm doing wrong?

GPU: MSI RX 480 8gb

RAM: 16GB Corsair Vengeance LPX @3000MHz and 1.35V

Cooler: Noctua-NHD15

Motherboard: Gigabyte Z170X-Designare

Are you monitoring your Vcore and Temps? If so, what vcore/temps are reported when you try a multiplier of 46?

 

[Makentox]

I can't get over 4.5 ghz with 1.35v on my CPU. It's not delidded or anything but this seems pretty bad. Did I just lose the silicon lottery or is there something I'm doing wrong?

GPU: MSI RX 480 8gb

RAM: 16GB Corsair Vengeance LPX @3000MHz and 1.35V

Cooler: Noctua-NHD15

Motherboard: Gigabyte Z170X-Designare

http://www.tomshardware.co.uk/answers/id-3313764/core-6700k.html?siteFrom=EPR-8807#xtor=EPR-8807

read last post for basic oc guide
dont mind my tragic english

 

[jbake47]

At stable 4.5 Ghz I see my Vcore spiking as high at 1.4v in both HWINFO and CPU-Z, when benching the temps max out at 80 C. When I run 4.6 it starts at 1.35v but as soon as I start Prime95 or Realbench it immediately gives me a blue screen, so I have no readings for 4.6.

 

[Makentox]

At stable 4.5 Ghz I see my Vcore spiking as high at 1.4v in both HWINFO and CPU-Z, when benching the temps max out at 80 C. When I run 4.6 it starts at 1.35v but as soon as I start Prime95 or Realbench it immediately gives me a blue screen, so I have no readings for 4.6.

never use auto voltage for overclocking

 

[burnhamjs]

At stable 4.5 Ghz I see my Vcore spiking as high at 1.4v in both HWINFO and CPU-Z, when benching the temps max out at 80 C. When I run 4.6 it starts at 1.35v but as soon as I start Prime95 or Realbench it immediately gives me a blue screen, so I have no readings for 4.6.

Yea, line Makentox says above you should manually set your vcore while determining your OC. (After you know the required vcore you can use an offset).

What is the max vcore you would be comfortable with? For me I'm not confortable going over 1.35v, some are ok going to 1.4v. I would manually set your vcore to the max you are comfortable with at 4.5GHz. Run stress/stability tests and see if you are stable and temps are <80C. If you are stable, try 4.6Ghz and monitor vcore/temps. If you are stable, trying to slowly back vcore down to the the minimum required to stay stable. If you are not stable at 4.6GHz, go back to 4.5G and slowly back vcore down to the the minimum required to stay stable.

 

[axlrose]

Haven't started overclocking yet, but plan to. Is 80c the magic ceiling for cpu's?

 

[burnhamjs]

Haven't started overclocking yet, but plan to. Is 80c the magic ceiling for cpu's?

During normal operations you would want temps to be much lower - but when you are OC'ing you will want to run thermal stress and stability testing. During those tests you want to keep temps below 80C (if they heat a peak just over 80C you would still be ok though).

 

[jbake47]

Hey guys what ended up being the limiting factor in my case was the CPU Loadline Calibration. By default it was set to Auto, but when I changed it to Turbo my voltages at higher clock speeds were a lot better! I was able to get 4.6 ghz stable at 1.35v(could probably lower the voltage) while overclocking my RAM to 3000 mhz 1.35v as well. Right now I'm running benchmarks on 4.7 ghz at 1.385v as well to see if I can get that stable as well.

 

[burnhamjs]

Hey guys what ended up being the limiting factor in my case was the CPU Loadline Calibration. By default it was set to Auto, but when I changed it to Turbo my voltages at higher clock speeds were a lot better! I was able to get 4.6 ghz stable at 1.35v(could probably lower the voltage) while overclocking my RAM to 3000 mhz 1.35v as well. Right now I'm running benchmarks on 4.7 ghz at 1.385v as well to see if I can get that stable as well.

Excellent. One thing you will want to keep an eye on is if that level of LLC (Load Line Calibration) causes Vboost.

With a manual vcore set, monitor your vcore in CPU-Z at idle and when under load (during stress/stability testing). If you see it drop significantly – that is Vdroop. If you see it ramp up significantly – that is Vboost. The level of LLC affects this. You want enough LLC that you don’t see a significant drop but not so high you see a significant boost. Basically, apply just enough LLC that the voltage stays relatively consistent at idle and under load (with a manual vcore).

If you see you are getting Vboost you could try to drop from Turbo to High.