Jump to content
HWBOT Community Forums

[Overclocking Support Thread] MSI X99 OC & GAMING Motherboards


mr. robot

Recommended Posts

Yes. Shut off and cold from the main switch. Reboots and warm shut downs not affected. Thanks.

 

 

Sent from my iPhone using Tapatalk

 

So, I've got a chat with the R&D guys about this double post behavior, see below. :)

 

1) Action: Turn PSU OFF and discharge the MB, then turn PSU and MB ON.

Behavior: POST will stop at code 33, shut down and reboot:

Explanation: This is related to the XMP LED (on the lower right corner of OC Genie button), the BIOS need to shut down to initiate the BIOS code related to the XMP LED.

 

2) Action: Clear CMOS

Behavior: MB will POST, then shut down before Windows loads and reboot

Explanation: This is for Intel ME to save current PCI-E lane switching and related configurations to ensure all PCI-E devices work well.

For e.g. without this mechanism, if all your PCI-E lanes are occupied and you clear CMOS + Remove some of the PCI-E devices, when you will turn ON the MB again the current PCI-E lane switching and related configurations will not be the same as the configurations saved in ME, which may cause issues with the PCI-E devices installed.

 

3) Action: Changing “the M.2 source link†between “auto†and ‘’PCH†in BIOS

Behavior: MB will POST, then shut down before Windows loads and reboot

Explanation: This is the same situation as 2.

 

More generally, if there is any change related to PCI-E, the double post behavior will happen for Intel ME to save the new configuration.

Edited by pepinorang
Link to comment
Share on other sites

  • Replies 439
  • Created
  • Last Reply

Top Posters In This Topic

About the Warning in a Test BIOS. I had seen it of course, but I for one would never bring out Test BIOS versions.

If MSI gives them to people you have to support it (Since it is for a MSI product).

 

About the first 5820K being crippled. It was with the X99S version of the XPower AC board.

I don't have it al with me here, but I had a problem running 2 SSD's (1 Win 7 and 1 Win 8) and I had a problem with the bios version I was running (I would have to look for the exact problem).

I made contact via the MSI website. I got a few beta BIOS versions. 1 of them killed the auto voltage selection part of my CPU. After that BIOS flash Auto selection for voltages wasn't working anymore.

I don't know how it was possible, I just noticed the changes. No update ever restored the functionality (By the way I sold it).

Link to comment
Share on other sites

First page updated with latest BIOS & Software versions. :celebration:

 

About the Warning in a Test BIOS. I had seen it of course, but I for one would never bring out Test BIOS versions.

If MSI gives them to people you have to support it (Since it is for a MSI product).

 

The TEST BIOS versions were requested by overclockers that work closely with us to test some OC features, ideas and/or remove some limitation, they were never meant to be released officially. After plenty of top scores were submitted with those BIOS, complains started to grow because 99% of the extreme overclockers here didn't have access to those "special" BIOS.

Thus, we decided once again to show our support to the HWBOT XOC community and did "a favor" to them by sharing those TEST BIOS on the forum.

Everything that was tested/unlocked in TC7/TC8 has been implemented into the OFFICIAL/BETA BIOS long time ago and bugs have been fixed as well. So in fact, that is the exact OPPOSITE of no vendor support! ;)

 

About the first 5820K being crippled. It was with the X99S version of the XPower AC board.

I don't have it al with me here, but I had a problem running 2 SSD's (1 Win 7 and 1 Win 8) and I had a problem with the bios version I was running (I would have to look for the exact problem).

I made contact via the MSI website. I got a few beta BIOS versions. 1 of them killed the auto voltage selection part of my CPU. After that BIOS flash Auto selection for voltages wasn't working anymore.

I don't know how it was possible, I just noticed the changes. No update ever restored the functionality (By the way I sold it).

 

"Auto selection of voltages wasn't working anymore" What do you mean exactly, in details?

Anyway, there is no such thing as an "auto voltage selection part" in any CPU.

The auto voltages are controlled by the BIOS, not by any part of the CPU.

Edited by pepinorang
Link to comment
Share on other sites

So, I've got a chat with the R&D guys about this double post behavior, see below. :)

 

 

 

1) Action: Turn PSU OFF and discharge the MB, then turn PSU and MB ON.

 

Behavior: POST will stop at code 33, shut down and reboot:

 

Explanation: This is related to the XMP LED (on the lower right corner of OC Genie button), the BIOS need to shut down to initiate the BIOS code related to the XMP LED.

 

 

 

2) Action: Clear CMOS

 

Behavior: MB will POST, then shut down before Windows loads and reboot

 

Explanation: This is for Intel ME to save current PCI-E lane switching and related configurations to ensure all PCI-E devices work well.

 

For e.g. without this mechanism, if all your PCI-E lanes are occupied and you clear CMOS + Remove some of the PCI-E devices, when you will turn ON the MB again the current PCI-E lane switching and related configurations will not be the same as the configurations saved in ME, which may cause issues with the PCI-E devices installed.

 

 

 

3) Action: Changing “the M.2 source link†between “auto†and ‘’PCH†in BIOS

 

Behavior: MB will POST, then shut down before Windows loads and reboot

 

Explanation: This is the same situation as 2.

 

 

 

More generally, if there is any change related to PCI-E, the double post behavior will happen for Intel ME to save the new configuration.

 

 

 

Thanks for the explanation but I don't understand why it forgets these configurations at every shutdown? This is very annoying. Why did the X99S AcK gaming 9 board not behave this way? Shouldn't the board save these configurations in cmos once and if no changes has occurred not go through this routine? Also have concerns what this shut down and power up again behavior will do or stress out my water pump?

 

Thanks.

Link to comment
Share on other sites

"Auto selection of voltages wasn't working anymore" What do you mean exactly, in details?

Anyway, there is no such thing as an "auto voltage selection part" in any CPU.

The auto voltages are controlled by the BIOS, not by any part of the CPU.

 

When I loaded bios defaults and I booted the pc it did actually start.

But when going to Windows (tested it with different images and Windows versions) it always crashed. After a while I discovered that when filling in the exact same (stock) voltage the cpu should be running on (from what I had from earlier testing) it actually worked fine. So it is something with voltages I always thought. Auto settings did not work for it.

 

Edit. Some writing went wrong. Please comment me for more.

Edited by Sparks.nl
Link to comment
Share on other sites

OK didnt ever believe this is possible but it seems one of the bios updates after 1.1 was killing my ram. Fact is i had massive boot problems and took out the fourth ram stick. And now everything is back normal and working like at the first day.

 

Tested many combinations but its the one ram stick for sure. Crazy now i need to RMA and so on and just on 12GB RAM. Maybe i just buy a new kit. Somebody have experiance with G-Skill RMA quality?

 

Whats up with bios 1.50 beta ?

Link to comment
Share on other sites

Ok flashed bios 1.51 yesterday and for me it seems to bring back some "Godlikeness". I double flashed bios and did CMOS reset. After that i put all my 4 ram sticks back, started and configured the bios. So from first time on a clean boot with m.2 and and XMP enabled also with all 4 sticks. No more "OC failed press F1" message or continous 04,60,55 debug LED boot loop. From that point good job!

 

For me only one problem persist now which is that my SteelSeries Sensei mouse will not get recognized after booting up the computer after a longer period in standby.

Link to comment
Share on other sites

Ok flashed bios 1.51 yesterday and for me it seems to bring back some "Godlikeness". I double flashed bios and did CMOS reset. After that i put all my 4 ram sticks back, started and configured the bios. So from first time on a clean boot with m.2 and and XMP enabled also with all 4 sticks. No more "OC failed press F1" message or continous 04,60,55 debug LED boot loop. From that point good job!

 

For me only one problem persist now which is that my SteelSeries Sensei mouse will not get recognized after booting up the computer after a longer period in standby.

 

 

 

What do you mean 'double flash'? Also does it fix the double post issue? Where do you dl it from?

 

Thanks

Link to comment
Share on other sites

What do you mean 'double flash'? Also does it fix the double post issue? Where do you dl it from?

 

Thanks

 

I mean with "double flash" that i flashed it one time ( that will update the updater aswell) and then after that i did flash it again and then shutdown and CMOS reset ( after disconnecting power).

 

It seems like it kinda fixed the double post from cold boot issue but not the intended "if u change pcie settings ME firmware needs to save and reboot" behaviour which seems a feature rather than a bug. Which to me is logical and i only have this if i am change critical values in the bios.

 

I did download it here in the frontpage.

Link to comment
Share on other sites

1)Thanks for the explanation but I don't understand why it forgets these configurations at every shutdown? 2) This is very annoying. 3) Why did the X99S AcK gaming 9 board not behave this way? Shouldn't the board save these configurations in cmos once and if no changes has occurred not go through this routine? Also have concerns what this shut down and power up again behavior will do or stress out my water pump?

 

Thanks.

 

1) You misread my explanation, Intel ME don't forget these configuration, that's the opposite. Because the saved configurations inside Intel ME might need to be updated, it double post to reset/update/save them.

2) You are not supposed to turn OFF your PSU each time your turn OFF your computer

3) Old ME versions didn't have this behavior but with newer Intel ME, ALL the X99/Z170 motherboards share this behavior.

4) See 1).

 

When I loaded bios defaults and I booted the pc it did actually start.

But when going to Windows (tested it with different images and Windows versions) it always crashed. After a while I discovered that when filling in the exact same (stock) voltage the cpu should be running on (from what I had from earlier testing) it actually worked fine. So it is something with voltages I always thought. Auto settings did not work for it.

 

Edit. Some writing went wrong. Please comment me for more.

 

Seems more like a BIOS issue... Anyway as said previously it's the BIOS/MB that controls the voltages (AUTO or manual), not the CPU.

 

Whats the gig with the 1.50 beta for godlike?

Any godlike features?

 

The full change log is always confidential, but for this one most change are related to Broadwell-E and bug fix. Here is what I can list:

- Fix system OC fail issue when use Avexir DDR4-3000 CL17 8GB memory and XMP is Enabled.

- Modify auto adjust DRAM timing related code.

- Modify auto adjust CPU Base Clock related code when use over XMP 3300MHz memory for Haswell-E CPU.

- Fix Intel SSD 750 S3 issue.

 

Hey Pepino,

 

You think those R&D guys can engineer the next bios release not to do a cmos shut down at XMP or M.2 PCI-E turning on?

 

Thanks

 

LOL, all I can say is re-read my explanation...

It's not possible and I already explained you why. ;)

 

OK didnt ever believe this is possible but it seems one of the bios updates after 1.1 was killing my ram. Fact is i had massive boot problems and took out the fourth ram stick. And now everything is back normal and working like at the first day.

 

Tested many combinations but its the one ram stick for sure. Crazy now i need to RMA and so on and just on 12GB RAM. Maybe i just buy a new kit. Somebody have experiance with G-Skill RMA quality?

 

Whats up with bios 1.50 beta ?

 

Ok flashed bios 1.51 yesterday and for me it seems to bring back some "Godlikeness". I double flashed bios and did CMOS reset. After that i put all my 4 ram sticks back, started and configured the bios. So from first time on a clean boot with m.2 and and XMP enabled also with all 4 sticks. No more "OC failed press F1" message or continous 04,60,55 debug LED boot loop. From that point good job!

 

For me only one problem persist now which is that my SteelSeries Sensei mouse will not get recognized after booting up the computer after a longer period in standby.

 

Double flash is 200% useless unless Intel ME update went wrong or it has been corrupted with BCLK OC, in which case any BCLK change would fail or result in 100MHz.

Since your memory hasn't been killed by "one of the bios updates after 1.1", which as I explained in a previous post with technical facts to someone else is IMPOSSIBLE; it's probably time to edit or delete your previous post, don't you think? :rolleyes:

Anyway, if I was you I would still try to run MEMTEST with all modules installed/detected at XMP, just to be sure as you may have one or more defective IC that cripple your kit.

Higher capacity/frequency kits are more affected by this than others.

Edited by pepinorang
Link to comment
Share on other sites

1) You misread my explanation, Intel ME don't forget these configuration, that's the opposite. Because the saved configurations inside Intel ME might need to be updated, it double post to reset/update/save them.

2) You are not supposed to turn OFF your PSU each time your turn OFF your computer

3) Old ME versions didn't have this behavior but with newer Intel ME, ALL the X99/Z170 motherboards share this behavior.

4) See 1).

 

 

 

1. So at every cold boot it needs to update configurations? Why doesnt the Gaming 9 X99S board act that way? and why is it limited to only when xmp or m.2 pcie is enabled?

 

2. I shut off the switch on back of the psu after computer shuts down everytime. Been doing it that way for years. Makes me feel safer.

 

3. The gaming 9 x99s has just as much of a recent ME than the godlike. Not only that my Asus Z170 Workstation doesnt doublepost either.

 

As far as Im conerned I can turn off XMP and manually set parameters if I have to. To me it sounds like a bug. Wish MSI fix it already.

 

Thanks.

Link to comment
Share on other sites

1. So at every cold boot it needs to update configurations? Why doesnt the Gaming 9 X99S board act that way? and why is it limited to only when xmp or m.2 pcie is enabled?

 

2. I shut off the switch on back of the psu after computer shuts down everytime. Been doing it that way for years. Makes me feel safer.

 

3. The gaming 9 x99s has just as much of a recent ME than the godlike. Not only that my Asus Z170 Workstation doesnt doublepost either.

 

Seriously, you are just asking the same questions again and again in a different way.

I already answered all with technical facts, one last time:

 

1. Please re-read previous post below, it has already answerer to your questions.

All X99/Z170 motherboards behave the exact same way besides only MSI latest models have XMP LED.

 

 

1) Action: Turn PSU OFF and discharge the MB, then turn PSU and MB ON.

Behavior: POST will stop at code 33, shut down and reboot:

Explanation: This is related to the XMP LED (on the lower right corner of OC Genie button), the BIOS need to shut down to initiate the BIOS code related to the XMP LED.

 

2) Action: Clear CMOS

Behavior: MB will POST, then shut down before Windows loads and reboot

Explanation: This is for Intel ME to save current PCI-E lane switching and related configurations to ensure all PCI-E devices work well.

For e.g. without this mechanism, if all your PCI-E lanes are occupied and you clear CMOS + Remove some of the PCI-E devices, when you will turn ON the MB again the current PCI-E lane switching and related configurations will not be the same as the configurations saved in ME, which may cause issues with the PCI-E devices installed.

 

3) Action: Changing “the M.2 source link†between “auto†and ‘’PCH†in BIOS

Behavior: MB will POST, then shut down before Windows loads and reboot

Explanation: This is the same situation as 2.

 

More generally, if there is any change related to PCI-E, the double post behavior will happen for Intel ME to save the new configuration.

 

2. That's up to you but it's not the way it's meant to be.

 

3. See 1.

 

As far as Im conerned I can turn off XMP and manually set parameters if I have to. To me it sounds like a bug. Wish MSI fix it already.

 

Thanks.

 

Or you can just leave your PSU ON like 99.9% of users out there...

I already explained to you with technical facts from our engineers that it's not a bug, but you still prefer to assume you are right? OK...

Edited by pepinorang
Link to comment
Share on other sites

Pepinorang,

 

My 5960 worked fine on the some motherboard and new bios versions did not correct it.

That would rule a bios mistake out.

 

So the CPU wasn't damaged by the board if it was working fine other motherboards... :rolleyes:

If BIOS wasn't the cause, you probably had some issue with MB voltage controller, either corrupted firmware or defective IC.

Link to comment
Share on other sites

Seriously, you are just asking the same questions again and again in a different way.

 

I already answered all with technical facts, one last time:

 

 

 

1. Please re-read previous post below, it has already answerer to your questions.

 

All X99/Z170 motherboards behave the exact same way besides only MSI latest models have XMP LED.

 

 

 

 

 

 

 

2. That's up to you but it's not the way it's meant to be.

 

 

 

3. See 1.

 

 

 

 

 

 

 

Or you can just leave your PSU ON like 99.9% of users out there...

 

I already explained to you with technical facts from our engineers that it's not a bug, but you still prefer to assume you are right? OK...

 

 

 

I'm sorry I'm not totally convinced so consider this a feature request if the bios R&D guys are listening. Thanks.

Link to comment
Share on other sites

1) You misread my explanation, Intel ME don't forget these configuration, that's the opposite. Because the saved configurations inside Intel ME might need to be updated, it double post to reset/update/save them.

2) You are not supposed to turn OFF your PSU each time your turn OFF your computer

3) Old ME versions didn't have this behavior but with newer Intel ME, ALL the X99/Z170 motherboards share this behavior.

4) See 1).

 

 

 

Seems more like a BIOS issue... Anyway as said previously it's the BIOS/MB that controls the voltages (AUTO or manual), not the CPU.

 

 

 

The full change log is always confidential, but for this one most change are related to Broadwell-E and bug fix. Here is what I can list:

- Fix system OC fail issue when use Avexir DDR4-3000 CL17 8GB memory and XMP is Enabled.

- Modify auto adjust DRAM timing related code.

- Modify auto adjust CPU Base Clock related code when use over XMP 3300MHz memory for Haswell-E CPU.

- Fix Intel SSD 750 S3 issue.

 

 

 

LOL, all I can say is re-read my explanation...

It's not possible and I already explained you why. ;)

 

 

 

 

 

Double flash is 200% useless unless Intel ME update went wrong or it has been corrupted with BCLK OC, in which case any BCLK change would fail or result in 100MHz.

Since your memory hasn't been killed by "one of the bios updates after 1.1", which as I explained in a previous post with technical facts to someone else is IMPOSSIBLE; it's probably time to edit or delete your previous post, don't you think? :rolleyes:

Anyway, if I was you I would still try to run MEMTEST with all modules installed/detected at XMP, just to be sure as you may have one or more defective IC that cripple your kit.

Higher capacity/frequency kits are more affected by this than others.

 

Haha ok thats not possible ??? Why the hell then memtest say my 4. ram is wrecked and with every 16gb kit it works? Why should the ic be crippled?? Because i just used XMP and everything other normal? Thats ridiculous especially in regards with the latest voltages mess in the 1.2 and 1.3 bios and a pulled of 1.3 bios tells me everything. Also the ram worked like a charm for many months with my settings without any issue until an bios update. I recheck it but in 99.9% its not the case here. With every other ram stick populated in the slot or in all slots its works just fine like before.

 

I am engineer at IBM so i dont edit my post cause its truly possible!!!

Edited by ChriscomIT
Link to comment
Share on other sites

1. So at every cold boot it needs to update configurations? Why doesnt the Gaming 9 X99S board act that way? and why is it limited to only when xmp or m.2 pcie is enabled?

 

2. I shut off the switch on back of the psu after computer shuts down everytime. Been doing it that way for years. Makes me feel safer.

 

3. The gaming 9 x99s has just as much of a recent ME than the godlike. Not only that my Asus Z170 Workstation doesnt doublepost either.

 

As far as Im conerned I can turn off XMP and manually set parameters if I have to. To me it sounds like a bug. Wish MSI fix it already.

 

Thanks.

 

When i got that right its mainly caused by XMP code getting enabled so turning XMP of and set it manally should eventually help u.

Link to comment
Share on other sites

So pepinorang u say its more likely that the 3000Mhz CL15 settings after around 5 months killed one of my IMCs (5930k IMC should be not that weak...) ? Randomly ? Short after a bios update. Let me recheck this. But this seems all very fishy. Regarding this there are also plenty threads about on the MSI forums where other people complaining about the same. I just say 04, 60, 55 debug code reboot loop with press F1 message cause OC failed. This caused by one ram stick. Without system runs fine.

Edited by ChriscomIT
Link to comment
Share on other sites

So the CPU wasn't damaged by the board if it was working fine other motherboards... :rolleyes:

If BIOS wasn't the cause, you probably had some issue with MB voltage controller, either corrupted firmware or defective IC.

 

No the cpu has the same problem on 2 other boards (still has, 1 year after the issues started).

It is just so that the 5960 works on the boards where the 5820 does not.

But a defective ic could be it.

Link to comment
Share on other sites

Haha ok thats not possible ??? Why the hell then memtest say my 4. ram is wrecked and with every 16gb kit it works? Why should the ic be crippled?? Because i just used XMP and everything other normal? Thats ridiculous especially in regards with the latest voltages mess in the 1.2 and 1.3 bios and a pulled of 1.3 bios tells me everything. Also the ram worked like a charm for many months with my settings without any issue until an bios update. I recheck it but in 99.9% its not the case here. With every other ram stick populated in the slot or in all slots its works just fine like before.

 

I am engineer at IBM so i dont edit my post cause its truly possible!!!

 

What voltage mess are you talking about? Any proofs? (screenshot or video)

I already explained in a previous post that the BIOS was removed because of Broadwell-E, you should ALL stop making baseless assumptions pulled out of thin air!

 

It's your argumentation that is hilarious, no technical facts, no proofs? It's just possible because you are "engineer at IBM"? AWESOME! :rolleyes:

 

So pepinorang u say its more likely that the 3000Mhz CL15 settings after around 5 months killed one of my IMCs (5930k IMC should be not that weak...) ? Randomly ? Short after a bios update. Let me recheck this. But this seems all very fishy. Regarding this there are also plenty threads about on the MSI forums where other people complaining about the same. I just say 04, 60, 55 debug code reboot loop with press F1 message cause OC failed. This caused by one ram stick. Without system runs fine.

 

Re-read my post, I never talked about IMC but IC (aka memory chips), I though an "engineer at IBM" would know the difference.

Show the threads, provide me proofs and technical facts, then I'll look into it...

So far it's just a bunch of hearsay from users with limited technical knowledge that still assume they know it all.

Edited by pepinorang
Link to comment
Share on other sites

No the cpu has the same problem on 2 other boards (still has, 1 year after the issues started).

It is just so that the 5960 works on the boards where the 5820 does not.

But a defective ic could be it.

 

Oh ok my bad, I was mistaking between the 5960 and the 5820.

So indeed, it was CPU related, however why not RMA it to Intel?

If as you say this was caused by a BETA BIOS update provided by FAE, why not ask MSI to replace it then?

Could you send me the Support ticket number that was provided to you and the FAE contact you dealed with in MP?

 

I'd like to investigate on that, thanks!

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...