Trouble Overclocking with MSI Z170A Gaming M5 + Intel 6600K Processor

Hey Wingman,

Sorry for the belated reply.

I noticed, looking at VID either in AIDA or HWmoniter that it stays static (Override Offset+). If I set things to Adaptive will I see the VID move up past 1.2198 (my usual VID at 4.0GHz)? When in Override mode it is rock steady, doesn't go up during load (which I believe it did when I had things set to Adaptive) nor does it go down. My cores are set to Fixed, not Dynamic so the Turbo/Speed step features are off by default. What I'm looking for is the VID to rise under load and to fall at idle, i.e., the most ideal conditions for chip longevity. Also, C-State doesn't seem to affect stability at all, regardless if my processor is set to Dynamic or Fixed. I believe it is part of the power-saving features, right?

-laz.

The terms "dynamic" and "adaptive" would seem to apply to your goals in that regard. Trouble is those terms will not necessarily be applied the same way by every manufacturer in how frequencies and voltages are affected. You will just have to experiment or wait for someone to reply who has the board or a similar one. Things may not even be the same in that respect from one board to another made by the same company, but you would think there would be some consistency, at least if in the same generation of technology.

Laz322 said:

Hey Wingman,

Sorry for the belated reply.

I noticed, looking at VID either in AIDA or HWmoniter that it stays static (Override Offset+). If I set things to Adaptive will I see the VID move up past 1.2198 (my usual VID at 4.0GHz)? When in Override mode it is rock steady, doesn't go up during load (which I believe it did when I had things set to Adaptive) nor does it go down. My cores are set to Fixed, not Dynamic so the Turbo/Speed step features are off by default. What I'm looking for is the VID to rise under load and to fall at idle, i.e., the most ideal conditions for chip longevity. Also, C-State doesn't seem to affect stability at all, regardless if my processor is set to Dynamic or Fixed. I believe it is part of the power-saving features, right?

-laz.

Click to expand...

When the mode is set to fixed voltage the voltage will not go up or down according to load and VID (Voltage identification), so VID is not used at all. When dynamic or adaptive Vcore is set the voltage will go up or down according to load using the VID to calibrate and it is a power saving feature that started in 2006. When you see the + core voltage symbol that increase the core voltage + VID. So it looks like this VID 1.2198 +0.075=1.2498

Turbo is the Clock speed boost for 2 cores 1 core when the processor is used in stock default configuration. Speed step is the power saving feature for clock speed on all cores and will be used with Dynamic Vcore also default Bios setting.

The easiest thing you can do is set your Bios to default then just change the multiplier then set the Dynamic Vcore to what you want for voltage under stress testing. It will work the way you want, clock speed and voltage will go up and down
with load or idle.

If your only going to overclock to 4.0GHz set the bios to default and just up the multiplier to 4.0GHz the processor probably won't need anymore voltage.

Post back what the MSI board has for Dynamic and Adaptive voltage settings?

Hey Wingman (Trents)

I'm just using 4.0GHz to measure against a semi-stable 4.3GHz overclock I did. I'm wondering if I, the human being, can tell the difference between these two different speeds. I know program monitoring/benching software can tell the difference, but can I? I'm also using it because it makes me feel better to see lower voltages, even though Intel has said voltages as high as 1.45v are fine for a Skylake 6600K (and 6700K/6800K?). Intel took the voltage as high as 1.52v, but did not recommend that end users ever attempt to do the same thing as it could lead to a damaged or fried processor.

I have to do more experimenting with Dynamic Vcore. I'm curious, when the cores are set to Dynamic do they all work (all four cores) at the highest setting (4.3 or whatever is the highest core) under load? And does the Vcore default to the highest voltage setting the user has set or does this get set automatically, according to Intel specs?

Thanks guys,

-Laz.

Laz322 said:

Hey Wingman (Trents)

I'm just using 4.0GHz to measure against a semi-stable 4.3GHz overclock I did. I'm wondering if I, the human being, can tell the difference between these two different speeds. I know program monitoring/benching software can tell the difference, but can I? I'm also using it because it makes me feel better to see lower voltages, even though Intel has said voltages as high as 1.45v are fine for a Skylake 6600K (and 6700K/6800K?). Intel took the voltage as high as 1.52v, but did not recommend that end users ever attempt to do the same thing as it could lead to a damaged or fried processor.

I have to do more experimenting with Dynamic Vcore. I'm curious, when the cores are set to Dynamic do they all work (all four cores) at the highest setting (4.3 or whatever is the highest core) under load? And does the Vcore default to the highest voltage setting the user has set or does this get set automatically, according to Intel specs?

Thanks guys,

-Laz.

Click to expand...

Going from 3.6GHz stock to 4.3GHz is more noticeable in programs then going from 4.0Ghz to 4.3GHz, With increasing from 3.5GHz to 4.0GHz there is about 10FPS increase.
Techspot LINK: http://www.techspot.com/review/1263-gears-of-war-4-benchmarks/page4.html



Do you have a link where Intel says that Core voltage 1.45v is fine for a i5 6600k?

With a overclock all cores will work at the same clock speed and Dynamic Vcore, going up and down according to load or idle.

wingman99 said:

Do you have a link where Intel says that Core voltage 1.45v is fine for a i5 6600k?

Click to expand...

I hope that was a typo from the OP @ 1.45v for a 6600K

Nope, that was not a typo at all. Intel has done internal testing at 1.520v volts and recommends users do not surpass the SAFE limit of 1.450v. Just Google Intel voltage specs for the 6600k. I think I have the Intel document somewhere on my system. I'll try and dig it up for you guys.

-laz.

Laz322 said:

Nope, that was not a typo at all. Intel has done internal testing at 1.520v volts and recommends users do not surpass the SAFE limit of 1.450v. Just Google Intel voltage specs for the 6600k. I think I have the Intel document somewhere on my system. I'll try and dig it up for you guys.

-laz.

Click to expand...

I would like to see that document, I have googled now and in the past and nothing.

Intel Skylake Data Sheet

wingman99 said:

I would like to see that document, I have googled now and in the past and nothing.

Click to expand...

Here's a link to the Intel document: 6th Generation Intel Core Processor Family Data Sheet

Page 113 out of 126 show the max voltage.

And here also Skylake Overclocking Guide

Here's an eTechnix overclock experiment

And an example of an end user going to 1.410v on an MSI board.

There is a lot of information and experimentation going on out there right now, average safe operating voltage max at 1.450v. Again, Intel has taken these chips to 1.520v.

Personally, I'd like to get a good overclock going within 1.25v-1.35v but would be willing to go as high as 1.40v (will better cooling) since the average end user's voltages are anywhere between 1.250v-1.450v. (bearing Intel's 1.520v and their recommendation not to exceed 1.450v) at speeds of 4.3GHz-4.8GHz.

-Laz.

Thanks for the links. I have seen the Intel 6th generation data sheet however that voltage is for stock clocks. After I called Intel they said they don't validate voltage plus increasing clock speed, Intel 1-916-377-7000. I was hoping you had some secret information from Intel, not the small sample of the overclocking community.

Overclocking is risky when spoke to Intel north America RMA/tech he said they have 3 a day with failure.

For 4.5GHz I have Vcore Maximum of 1.339v.

wingman99 said:

Thanks for the links. I have seen the Intel 6th generation data sheet however that voltage is for stock clocks. After I called Intel they said they don't validate voltage plus increasing clock speed, Intel 1-916-377-7000. I was hoping you had some secret information from Intel, not the small sample of the overclocking community.

Overclocking is risky when spoke to Intel north America RMA/tech he said they have 3 a day with failure.

For 4.5GHz I have Vcore Maximum of 1.339v.

Click to expand...

I did read something that was on the line of Intel giving the go-ahead for anything not going past 1.45v. I'll try and find it for you. Again, Intel really doesn't get into telling their customers what's safe or not safe when it comes to overclocking, even though they market the K series chips for overclocking. I guess it is a legal issue or something.

As for sample-size, I could provide literally dozens and dozens of people (half of them professionals with tech sites and YouTube pages, etc.) overclocking these chips a 1.45v or greater without problems. That doesn't necessarily mean I want to follow in their footsteps. I'm a bit on the conservative side just like you. Also, since these chips are new we have no idea as to their longevity at higher then recommended voltages. They might last for ten years or they might die in five, we just don't know yet.

Being in the forums and talking to Intel RMA/Tech, I see dozens of K CPU's that fail on the average of 6 months overclocked. So I don't like to be to risky because my CPU could be just in specification to run stock and If pushed to far it will fail.

Hey wingman,

I'm sure, that's because people are doing things they shouldn't, without knowing what specific Bios settings do or applying way too much voltage to their chip. That and they're probably a lot of young kids making mistakes, right along with some of us adults. Trust me, I'm not too keen on upping my Vcore voltage unless I absolutely have to in order to create a stable overclock.

What's bugging me is that my results are no where near what seems to be the common results people are getting with this chip *6600K*, in regard to voltage and clock speeds. I must have read at least 20 articles or forum posts where people are getting to 4.4GHz easy, and with a very modest voltage increase (1.25v-1.28v). I haven't been able to get 4.3GHz stable with 1.28v or even 1.30v!

From what I've gathered my mobo (MSI Z1780A Gaming M5) isn't all that great for overclocking, even though it is marketed as a "Gaming/Overclocking" motherboard. For starters it doesn't have LLC in the Bios and MSI tech support told me that it will never be added. Why? Because it has to be done on the hardware side of the motherboard, meaning that the circuitry itself is the limiting factor, not code that can be added to the Bios. When I heard this news I was surprised and disappointed. I sincerely thought that LLC could be added via a Bios update. I'm still not sure I believe MSI. Even budget boards usually have LLC (Load Line Calibration).

So why didn't MSI add LLC to a board they claim was designed for overclocking?


A. Because the new Skylake chips make overclocking easier and voltages aren't as important with Skylake or B. The want people to buy the more expensive M7 motherboard just to have LLC


To be honest, neither of those reasons make much sense to me. MSI doesn't even mention LLC in their advertisements for either board. It's something the end user would find out about after buying the board, much to his or her dismay, at least if they planned to overclock and were unfortunate enough to by the M5. As for the other reason why MSI didn't include LLC on the M5, that doesn't make sense either since a lot of the controls that were traditionally left to the processor have now been switched to the motherboard (meaning they are adjusted through the motherboard whereas before they were adjusted by the processor itself).

Every review I read about this board before I actually bought it made the MSI M5 out to be a really solid board for overclocking, every damn review. Those reviews led me to buying this board, specifically because of the overclocking ability of the board. So I buy it, install it and begin to overclock and can't get anywhere near the speeds and voltages people got in those reviews and in their YouTube videos.

I checked out the settings in the Bios that MSI used for their "Game Boost" software overclock. The voltages for the MSI Game Boost of 4.1GHZ were 1.41v!!!

Sounds really high to me, especially for only 4.1GHz. Compare that to the numbers floating around everywhere else and you'll quickly realize that something must be wrong, somewhere, whether it's the processor itself or the board. Personally, I think the board is the limiting factor but I won't know until I try another motherboard or through an identical processor on this board and see how it overclocks.

The final possibility is a setting in the Bios that is somehow limiting my overclock and causing me to apply more voltage then what one would consider normal. I'm going to keep trying and see if I can somehow reach 4.4GHZ without going past 1.35v, that at least, should be possible.

-Laz.

Folks always say they must be doing something wrong when they see overclocking CPU failure. However the cases I see in the forums this year for Skylake they are always using low voltage so they say.
How can you blame folks anyway you are having a hard time understand Dynaimic Vcore, LLC, and how unimportant a motherboard is for a mid overclock with skylake.

LLC (Load line Calibration) has nothing to do with overclocking, folks can still achieve the same Vcore voltage with it not being used. LLC is for bypassing Intel's specification for Vdroop under load. All you have to do is run Adaptive + Offset Vcore and the peak voltage will fall and rise with CPU load variation.

With Load Line Calibration disabled in BIOS, setting a CPU Voltage VID of 1.38750 resulted in a no-load voltage of about 1.34V and a full-load value of 1.28V. Enabling this feature and lowering the VID to 1.35000V produced a constant CPU supply voltage, regardless of load (or so it seemed), of 1.33V. Setting a lower VID resulted in a blue screen during Windows boot. Idle voltage was relatively unchanged at about 1.33-1.34V but the full-load voltage required increased by 50mV with no benefit. As you might guess, we recommend you leave this option disabled.

Hopefully we've shown you enough to understand exactly why Voffset and Vdroop are important. Please give second thought to your actions if you're in the habit of defeating these essential system safeguards. http://www.anandtech.com/show/2404/6

Click to expand...

It is not your motherboard or any motherboard for a mid overclock on skylake, they all do just fine. All the motherboard has to do is supply more voltage+AMPs to the CPU for the slight increase in clock speed and voltage. 1.25v-1.28v is stock voltage for i7 6700k at 4.0GHz not i5 6600k.

As far as what you see for voltages you can't trust what you see in forums. Example with my i5 6600k the voltage for 4.5GHz is 1.272v to 1.339v depending on what I stress test. I see a lot of complaints just like yours where they need higher voltage than mine for 4.4 to 4.5GHz

So to sum it up you lost the luck of the draw in CPUs. I have had bad luck my self many times, now I have CPU that clocks to 4.5GHz with ease. I have been doing this since 2000 and there is more bad luck than good luck with overclocking.

wingman99 said:

How can you blame folks anyway you are having a hard time understand Dynaimic Vcore, LLC, and how unimportant a motherboard is for a mid overclock with skylake.

LLC (Load line Calibration) has nothing to do with overclocking, folks can still achieve the same Vcore voltage with it not being used. LLC is for bypassing Intel's specification for Vdroop under load. All you have to do is run Adaptive + Offset Vcore and the peak voltage will fall and rise with CPU load variation.


It is not your motherboard or any motherboard for a mid overclock on skylake, they all do just fine. All the motherboard has to do is supply more voltage+AMPs to the CPU for the slight increase in clock speed and voltage. 1.25v-1.28v is stock voltage for i7 6700k at 4.0GHz not i5 6600k.

As far as what you see for voltages you can't trust what you see in forums. Example with my i5 6600k the voltage for 4.5GHz is 1.272v to 1.339v depending on what I stress test. I see a lot of complaints just like yours where they need higher voltage than mine for 4.4 to 4.5GHz

So to sum it up you lost the luck of the draw in CPUs. I have had bad luck my self many times, now I have CPU that clocks to 4.5GHz with ease. I have been doing this since 2000 and there is more bad luck than good luck with overclocking.

Click to expand...

No, I know how to use the different offsets as I've already dealt with Vdroop but the voltage is still high according to even the Silicon Lottery 80% value. You may be right, I may have lost the lottery but then again, it may very well be the board, regardless if you're unaware of the problems some people have been having. A friend of mine on another forum said he believed that using the offset settings was better then LLC, I just haven't made up my mind in that regard yet, he may be right also.

The only way to really know is to throw another 6600K into my machine and see how it handles. And no, I'm not having a hard time understanding the offset, it's very straight forward. LLC does the same thing for the most part but does it in a different way according to motherboard circuitry, or so I have been told by MSI.

It is not your motherboard or any motherboard for a mid overclock on skylake, they all do just fine

Click to expand...

You win the gross generalization award of the day, congratulations! You're contradicting yourself. If you want to help me out, fine, but please, don't insult me wingman.

-laz.

laz, I think wing is correct and I suggested that to you as well, that the chip is just not a good overclocker. It seems to fall in the left side of the overclocking potential bell curve. You seem to have tried everything else and understand the process. Yes, it is true, one should be able to accomplish with offset what LLC does anyway and you get that so not much else to do except try another chip.

Laz322 said:

No, I know how to use the different offsets as I've already dealt with Vdroop but the voltage is still high according to even the Silicon Lottery 80% value. You may be right, I may have lost the lottery but then again, it may very well be the board, regardless if you're unaware of the problems some people have been having. A friend of mine on another forum said he believed that using the offset settings was better then LLC, I just haven't made up my mind in that regard yet, he may be right also.

The only way to really know is to throw another 6600K into my machine and see how it handles. And no, I'm not having a hard time understanding the offset, it's very straight forward. LLC does the same thing for the most part but does it in a different way according to motherboard circuitry, or so I have been told by MSI.



You win the gross generalization award of the day, congratulations! You're contradicting yourself. If you want to help me out, fine, but please, don't insult me wingman.

-laz.

Click to expand...

I'm not trying to insult you, I know how LLC and VRM Buck converter work. So when I read what your are saying about LLC and Adaptive + Offset Vcore, also your motherboard, I can't figure out what you know.

I'm not contradicting my self, all Z170 motherboards are made for overclocking and capable of mid overclocking just fine, It is the processors that are not able to overclock the same or use the same voltages.

The problem with overclocking a lot folks don't actually know what the settings in Bios do, so the is a lot of misunderstanding and false information.

No, I know how to use the different offsets as I've already dealt with Vdroop but the voltage is still high according to even the Silicon Lottery 80% value.

Click to expand...

Silicon Lottery is 1.392V CPU VCORE (Or less) for 4.6GHz. I know you can use offsets, however do you know what they are doing?

A friend of mine on another forum said he believed that using the offset settings was better then LLC, I just haven't made up my mind in that regard yet, he may be right also.

Click to expand...

I will go over this again. Adaptive + Offset, means VID=(Adaptive), the VID is communicated by the microprocessor to the VRM + the offset voltage you set.

Separate CPU's in the same class have varying VID, so every CPU has it's own VID.

Override Mode = fixed voltage so it does not use the Dynamic VID.
Adaptive + Offset = VID + offset.
Override + Offset = Fixed voltage and Fixed voltage.

LLC (Load Line Calibration) is a Intel specification to prevent over shoot with voltage when there is a varying load.

Read this

If you've ever overclocked a system, chances are that at some point or another you've had opportunity to become upset with your Vdroop "problem." Some users, confused as to why their system refuses to exactly match actual processor supply voltage to the value specified in BIOS, are quick to blame the quality their motherboard; still others find fault with the difference noted between their board's idle and full-load processor supply voltages. Actually, load line droop (Vdroop) is an inherent part of any Intel power delivery design specification and serves an important role in maintaining system stability. In most cases, comments regarding unacceptable power delivery performance are completely unfounded. To make matters worse, unjustified negative consumer perception surrounding this often misunderstood design feature eventually forced a few motherboard manufacturers to respond to enthusiasts' demands for action by adding an option in their BIOS that effectively disables this important function. http://www.anandtech.com/show/2404/5

Click to expand...

You may be right, I may have lost the lottery but then again, it may very well be the board, regardless if you're unaware of the problems some people have been having.

Click to expand...

What problems are some people having with the MSI Z170A Gaming M5?

trents said:

laz, I think wing is correct and I suggested that to you as well, that the chip is just not a good overclocker. It seems to fall in the left side of the overclocking potential bell curve. You seem to have tried everything else and understand the process. Yes, it is true, one should be able to accomplish with offset what LLC does anyway and you get that so not much else to do except try another chip.

Click to expand...

Yeah. I did say wingman might be right but I'm just not ready to throw in the towel just yet. I do appreciate everyone's comments I just think we need to be careful in how we interact with one another and mind what we say. Thanks Trents.

Laz322 said:

Yeah. I did say wingman might be right but I'm just not ready to throw in the towel just yet. I do appreciate everyone's comments I just think we need to be careful in how we interact with one another and mind what we say. Thanks Trents.

Click to expand...

Yes, things can get a little chippy sometimes and courtesy is always in order. At times over the years on the forum I have gotten a little terse with some individuals so I'm preaching to myself here.

trents said:

Yes, things can get a little chippy sometimes and courtesy is always in order. At times over the years on the forum I have gotten a little terse with some individuals so I'm preaching to myself here.

Click to expand...

Definitely, and sometimes my skin can be a little on the thin side. I understand Wingman was just trying to help and I'm really not interested in arguing, though my ego sometimes overrides my common sense and peaceful nature (yeah right! )

-laz,