Jump to content

i5-6600K overclock questions

Am I doing it wrong or that cpu overclock pretty easily?

I went from 3.5 to 4.2GHz without even changing core voltage, which was set to 1.18V instead of 1.2V. I have yet to try 4.3GHz, but it seems so easy I'm wondering if I'm doing something wrong to be honest.

Also, putting half integer multiplier doesn't seem to work, I wonder if someone can explain me why you can't put 42.5, making a 4.25 GHz frequency?

Finally, should cache frequencies be changed or does it not matter?

Link to comment
Share on other sites

Link to post
Share on other sites

You went from 3.5 to 4.2 without changing voltage, probably because it's set on Auto by default. So it's just going to use whatever voltage it needs to maintain.

 

Who knows about the 42.5 thing, if you REALLY want that kinda OC then mess with the BCLK, 

 

Plus don't overclock on auto voltage, you could be running at 1.45+ voltages and really damaging your CPU. OC with manually set voltages.

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, Katsunaka said:

You went from 3.5 to 4.2 without changing voltage, probably because it's set on Auto by default. So it's just going to use whatever voltage it needs to maintain.

 

Who knows about the 42.5 thing, if you REALLY want that kinda OC then mess with the BCLK, 

 

Plus don't overclock on auto voltage, you could be running at 1.45+ voltages and really damaging your CPU. OC with manually set voltages.

Thanks I'll have to check, but I remember voltage being put at 1.18V though.

At that frequency I don't really think it should run at 1.45V though :)

 

I don't really want to, but I'm curious as to why

Link to comment
Share on other sites

Link to post
Share on other sites

7 minutes ago, laminutederire said:

Also, putting half integer multiplier doesn't seem to work, I wonder if someone can explain me why you can't put 42.5, making a 4.25 GHz frequency?

It's a clock multiplier, those just normally only work in integer multiples (there have been half-steps in some cases in the past, but not on any current hardware AFAIK). I couldn't tell you exactly why, but it does just seem like a simpler and easier operation.

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, laminutederire said:

At that frequency I don't really think it should run at 1.45V though :)

It's not a matter of "should it run", it's a matter of the mobo trying to "make sure" it works.

11 minutes ago, laminutederire said:

Also, putting half integer multiplier doesn't seem to work, I wonder if someone can explain me why you can't put 42.5, making a 4.25 GHz frequency?

Because multipliers are integer numbers. Simple as that. They don't accept floating point numbers because that's the way they were designed (working with integers is just soooooo much easier). That's not to say you can't have 4.25ghz, though, all you'd need to do is change the base clock such that Blck*Multiplier = 4.25ghz.

 

13 minutes ago, laminutederire said:

I went from 3.5 to 4.2GHz without even changing core voltage, which was set to 1.18V instead of 1.2V. I have yet to try 4.3GHz, but it seems so easy I'm wondering if I'm doing something wrong to be honest.

Did you stress test it? If so, how did you do it?

13 minutes ago, laminutederire said:

Finally, should cache frequencies be changed or does it not matter?

Depends on what you are doing. For most things, OCing cache doesn't help. You can mess with it later if you want, leave it at stock for now.

Want to help researchers improve the lives on millions of people with just your computer? Then join World Community Grid distributed computing, and start helping the world to solve it's most difficult problems!

 

Link to comment
Share on other sites

Link to post
Share on other sites

3 minutes ago, Sakkura said:

It's a clock multiplier, those just normally only work in integer multiples (there have been half-steps in some cases in the past, but not on any current hardware AFAIK). I couldn't tell you exactly why, but it does just seem like a simpler and easier operation.

Yeah or you change BCLK to half and you can achieve what you want :)

There are no reasons why not though

2 minutes ago, Imakuni said:

It's not a matter of "should it run", it's a matter of the mobo trying to "make sure" it works.

Because multipliers are integer numbers. Simple as that. They don't accept floating point numbers because that's the way they were designed (working with integers is just soooooo much easier). That's not to say you can't have 4.25ghz, though, all you'd need to do is change the base clock such that Blck*Multiplier = 4.25ghz.

 

Did you stress test it? If so, how did you do it?

Depends on what you are doing. For most things, OCing cache doesn't help. You can mess with it later if you want, leave it at stock for now.

The voltage was what I saw in the bios, but I ran another prime torture test, and indeed voltage goes higher (highest was 1.36V, which isn't good).

I'll finish installing something and I'm fiddling with that right after that though :/

Link to comment
Share on other sites

Link to post
Share on other sites

16 minutes ago, Katsunaka said:

You went from 3.5 to 4.2 without changing voltage, probably because it's set on Auto by default. So it's just going to use whatever voltage it needs to maintain.

 

Who knows about the 42.5 thing, if you REALLY want that kinda OC then mess with the BCLK, 

 

Plus don't overclock on auto voltage, you could be running at 1.45+ voltages and really damaging your CPU. OC with manually set voltages.

It wasn't on auto it was on offset mode from 1.184V.

Link to comment
Share on other sites

Link to post
Share on other sites

5 minutes ago, laminutederire said:

I ran another prime torture test

I'd love if you could be a bit more specific on that...

6 minutes ago, laminutederire said:

I'll finish installing something and I'm fiddling with that right after that though

FYI, mine stopped at 4.3ghz core at 1.3v. Good luck.

Want to help researchers improve the lives on millions of people with just your computer? Then join World Community Grid distributed computing, and start helping the world to solve it's most difficult problems!

 

Link to comment
Share on other sites

Link to post
Share on other sites

2 minutes ago, Imakuni said:

I'd love if you could be a bit more specific on that...

FYI, mine stopped at 4.3ghz core at 1.3v. Good luck.

Well I ran a prime95 torture test, and I checked only the Temps. 

I redid the whole thing to put it on manual with 1.2V, but in cpu-Z the core vid goes up to 1.35V still, is that normal? 

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, laminutederire said:

Well I ran a prime95 torture test, and I checked only the Temps.

Sigh... there are 3 default mode (+ the custom one). Which did you use? Also, which version of P95 was used (this matters A LOT)? Last, what temps did you get?

2 minutes ago, laminutederire said:

I redid the whole thing to put it on manual with 1.2V, but in cpu-Z the core vid goes up to 1.35V still, is that normal? 

Could you post a screenshot of HWMonitor please? Show the motherboard and CPU sections, the others aren't relevant.

Want to help researchers improve the lives on millions of people with just your computer? Then join World Community Grid distributed computing, and start helping the world to solve it's most difficult problems!

 

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, Imakuni said:

Sigh... there are 3 default mode (+ the custom one). Which did you use? Also, which version of P95 was used (this matters A LOT)? Last, what temps did you get?

Could you post a screenshot of HWMonitor please? Show the motherboard and CPU sections, the others aren't relevant.

oh that, sorry. I used small FFT's which got the core up to 65 degrees celsius, and I did a blend test as well which got me to less (upper 50's if I remember right)

I have HWinfo64 which should be equivalent.

Here is the cpu part:

f1i42cc.png

and that's the mobo part:

tlJPPVA.png

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, laminutederire said:

oh that, sorry. I used small FFT's which got the core up to 65 degrees celsius, and I did a blend test as well which got me to less (upper 50's if I remember right)

Huh. Are you using version 28.9?

2 minutes ago, laminutederire said:

Here is the cpu part:

Well, seems like it's going up to 1.35. You are using Fixed (not auto, not offset) voltage, right? If not, put it at FIxed. If so, seems like your LLC is poorly calibrated and you'll have to adjust it.

Want to help researchers improve the lives on millions of people with just your computer? Then join World Community Grid distributed computing, and start helping the world to solve it's most difficult problems!

 

Link to comment
Share on other sites

Link to post
Share on other sites

Just now, Imakuni said:

Huh. Are you using version 28.9?

Well, seems like it's going up to 1.35. You are using Fixed (not auto, not offset) voltage, right? If not, put it at FIxed. If so, seems like your LLC is poorly calibrated and you'll have to adjust it.

Yes I am.

Vcore on the mainboard is about right what it is supposed to. (in hwinfo that is) but core VID goes all over the place. It is supposed to be a fixed. There are four modes in BIOS: auto, manual, offset and adaptative (which is offset with a "turbo" mode.

I set it to manual.

LLC? seems awfully dangerous said like that ^^

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, laminutederire said:

Yes I am.

Vcore on the mainboard is about right what it is supposed to. (in hwinfo that is) but core VID goes all over the place. It is supposed to be a fixed. There are four modes in BIOS: auto, manual, offset and adaptative (which is offset with a "turbo" mode.

I set it to manual.

LLC? seems awfully dangerous said like that ^^

Load Line Calibration. Basically, it increases the voltage when the load is high, to make up for droops.

 

As an example, I've set my voltage to 1.27. If Iet LLC at Lvl 4, it droops to 1.24, so I've had to set it to level 3 to achieve 1.3. Lvl 2 would shoot me up to 1.33. Play around with it and see what works best for you.

Want to help researchers improve the lives on millions of people with just your computer? Then join World Community Grid distributed computing, and start helping the world to solve it's most difficult problems!

 

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, Imakuni said:

Load Line Calibration. Basically, it increases the voltage when the load is high, to make up for droops.

 

As an example, I've set my voltage to 1.27. If Iet LLC at Lvl 4, it droops to 1.24, so I've had to set it to level 3 to achieve 1.3. Lvl 2 would shoot me up to 1.33. Play around with it and see what works best for you.

from what I gathered about vid and vcore from the internet, it's what the core is demanding, while vcore is what's actually supplied, so if the vid goes all over the place it doesn't matter as long as the vcore doesn't go as high and stays at what it's supposed to?

Link to comment
Share on other sites

Link to post
Share on other sites

2 minutes ago, laminutederire said:

from what I gathered about vid and vcore from the internet, it's what the core is demanding, while vcore is what's actually supplied, so if the vid goes all over the place it doesn't matter as long as the vcore doesn't go as high and stays at what it's supposed to?

Skylake readings aren't exactly the most reliable thing. Just look at it: do you really think your CPU is only requesting 1.184v under high load? I've seen people in with bad readings on the mobo and bad readings on the CPU, but I've also seen good readings on both fronts.

 

In my specific case, readings from the CPU are wrong and mobo are right. In yours, looks like you should ignore the mobo readings and focus on the CPU ones.

Want to help researchers improve the lives on millions of people with just your computer? Then join World Community Grid distributed computing, and start helping the world to solve it's most difficult problems!

 

Link to comment
Share on other sites

Link to post
Share on other sites

Just now, Imakuni said:

Skylake readings aren't exactly the most reliable thing. Just look at it: do you really think your CPU is only requesting 1.184v under high load? I've seen people in with bad readings on the mobo and bad readings on the CPU, but I've also seen good readings on both fronts.

 

In my specific case, readings from the CPU are wrong and mobo are right. In yours, looks like you should ignore the mobo readings and focus on the CPU ones.

well it's what it's supposed to be limited at in the bios, so I hope the mobo readings are the right ones

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×