Boot "problems" with new cpu

[Ka0s]

Hi mates !

I finally chose a 5600x and i am very happy with the change. After mounting it, renewing the paste and formatting it works perfectly.

The only "problem" is that i must have some erroneous configuration that makes that when activating the DOCP mode of the memory for the 3200Mhz it does a strange thing. After a power loss (unplugged and plugged again), when booting, the fans turn on for 1 or 2 seconds, turn off, do the same again and the third time it starts correctly.

I have tried to load optimal values in the bios and the problem disappears.


Any tip?
Thanks!

-

Edit for system specs:

5600x
Asus Rog Strix x370-F Gaming
2 x 8GB GSkill 3200MHz CL14 BDie

 

[TheoneandonlyMrK]

Hi mates !

I finally chose a 5600x and i am very happy with the change. After mounting it, renewing the paste and formatting it works perfectly.

The only "problem" is that i must have some erroneous configuration that makes that when activating the DOCP mode of the memory for the 3200Mhz it does a strange thing. After a power loss (unplugged and plugged again), when booting, the fans turn on for 1 or 2 seconds, turn off, do the same again and the third time it starts correctly.

I have tried to load optimal values in the bios and the problem disappears.


Any tip?
Thanks!

It's due to memory training and is normal.
As you say if You set timings instead of it setting them it goes away.

 

[Ka0s]

I thought it might be the base clock, but no idea...

 

[TheoneandonlyMrK]

I thought it might be the base clock, but no idea...

Np definitely memory training, mine does the same.

 

[P4-630]

Np definitely memory training, mine does the same.

Some motherboards has options to (partially) turn that off iirc, however I don't have AMD myself though.

 

[TheoneandonlyMrK]

Some motherboards has options to (partially) turn that off iirc, however I don't have AMD myself though.

Most DOCP profiles set only the first four to six settings leaving the rest on auto, so memory training has to work to find them.

Every X370,470,550,570 board I have setup DOCP on has done this.

You can turn it off on every board by manual configuration of all the memory timings,or using Jdec defaults and that's all Afaik.

 

[dont whant to set it"']

@Ka0s There should be a BIOS/uefi setting somewhere about ram training reattempt count, default is set to 3, so that is why. Even if it trains successfully the first time , the mb will go trough the loop.
If WHEA errors or bsod, then some tunning may be need for something as simple as raising VSoC voltage by 15to25 miliVolts does the trick, not a rule.

 

[The King]

Post your Zentimings with XMP enabled. Note down what ProcODT is and set that manually in your BIOS.
This may help the board to boot or train faster.

 

[bug]

Hi mates !

I finally chose a 5600x and i am very happy with the change. After mounting it, renewing the paste and formatting it works perfectly.

The only "problem" is that i must have some erroneous configuration that makes that when activating the DOCP mode of the memory for the 3200Mhz it does a strange thing. After a power loss (unplugged and plugged again), when booting, the fans turn on for 1 or 2 seconds, turn off, do the same again and the third time it starts correctly.

I have tried to load optimal values in the bios and the problem disappears.


Any tip?
Thanks!

-

Edit for system specs:

5600x
Asus Rog Strix x370-F Gaming
2 x 8GB GSkill 3200MHz CL14 BDie

That usually happens when the system can't boot with the current settings. It reboots and resets BIOS settings. Check RAM timings after it does that, it may have reverted to non-DOCP.

 

[Ka0s]

Post your Zentimings with XMP enabled. Note down what ProcODT is and set that manually in your BIOS.
This may help the board to boot or train faster.

Here is it.
What is strange to me is that in the default mode (no docp/xmp active and everything automatic) it works perfectly.

The only change i made was to activate docp (3200mhz auto settings), which did not happen with the 1600x and with the 5600x it did.


timings.jpg >> Just DOCP/XMP active, no manual settings.
noxmp.jpg >> Without DOCP/XMP active, full auto and optimized default settings on Bios.

 

[TheoneandonlyMrK]

Here is it.
What is strange to me is that in the default mode (no docp/xmp active and everything automatic) it works perfectly.

The only change i made was to activate docp (3200mhz auto settings), which did not happen with the 1600x and with the 5600x it did.


timings.jpg >> Just DOCP/XMP active, no manual settings.
noxmp.jpg >> Without DOCP/XMP active, full auto and optimized default settings on Bios.

Read AMD's or your motherboard makers manual, I already told you what it says though.

Or ignore me and hunt for this non issues fix not fix some more but I'm out bye and good luck with your problem!?.

 

[Ka0s]

Read AMD's or your motherboard makers manual, I already told you what it says though.

Or ignore me and hunt for this non issues fix not fix some more but I'm out bye and good luck with your problem!?.

Sorry !, im just reply the post of The King user.
I read your post too mate, not ignoring you. Really sorry about missunderstanding

 

[Bill_Bright]

This is NOT normal!

When you first install the RAM (or make changes to the RAM configuration) "training" may be required on the first boot after the change is made. And the training process could take several minute and require yet another reboot. But once complete, it "should" boot normally after that (until another change).

If you have to go through the process of training over and over again, or the day after day boot process requires 3 times to finally boot into the OS, then something is not right! It could be incompatible RAM; is yours listed on the QVL here? Or it could be faulty RAM - have you tried just one stick at a time? Or it could be an unstable setting. Or it could be your CMOS settings are not holding properly.

You say this happens after a full power loss. What happens if you "gracefully" shutdown Windows and your computer normally? Does it boot in one attempt after that?

What happens if you shutdown gracefully, then flip the PSU's master power switch to "0" or "Off" (or unplug from the wall) for a minute or two, then restore power and boot?

Have you reset your BIOS and started over? If me, since they are inexpensive, I would replace the CMOS battery. While they should last many years, some don't. I have even seen DOA, brand new CR2032 batteries right out of the package.

 

[TheoneandonlyMrK]

Hi mates !

I finally chose a 5600x and i am very happy with the change. After mounting it, renewing the paste and formatting it works perfectly.

The only "problem" is that i must have some erroneous configuration that makes that when activating the DOCP mode of the memory for the 3200Mhz it does a strange thing. After a power loss (unplugged and plugged again), when booting, the fans turn on for 1 or 2 seconds, turn off, do the same again and the third time it starts correctly.

I have tried to load optimal values in the bios and the problem disappears.


Any tip?
Thanks!

-

Edit for system specs:

5600x
Asus Rog Strix x370-F Gaming
2 x 8GB GSkill 3200MHz CL14 BDie

Read the bold bit Bill, it's normal.

 

[The King]

Here is it.
What is strange to me is that in the default mode (no docp/xmp active and everything automatic) it works perfectly.

The only change i made was to activate docp (3200mhz auto settings), which did not happen with the 1600x and with the 5600x it did.


timings.jpg >> Just DOCP/XMP active, no manual settings.
noxmp.jpg >> Without DOCP/XMP active, full auto and optimized default settings on Bios.

60 ohms looks to be abit off to me its on the high side.
on ZEN 3 it should be around 36.9 Ohms. This is basically the same ram you have Samsung B-die.

 

[Bill_Bright]

Read the bold bit Bill, it's normal.

I did. And I say again, it is not normal.

Even after an unexpected power loss, it should not require multiple reboots. Unexpected power outages are never good, but the worst that "could" happen (under normal conditions) is a corrupt drive as opened files might not be closed properly. But the BIOS settings "should" not be affected. They "should" hold. That's exactly what the CMOS battery is there for (that, and to keep the RTC ticking).

 

[The King]

I did. And I say again, it is not normal.

Even after an unexpected power loss, it should not require multiple reboots. Unexpected power outages are never good, but the worst that "could" happen (under normal conditions) is a corrupt drive as opened files might not be closed properly. But the BIOS settings "should" not be affected. They "should" hold. That's exactly what the CMOS battery is there for (that, and to keep the RTC ticking).

This is correct. Once the board trains it should not do it again after a reboot or from a cold restart. Since he has an older X370 mobo it is possible his CMOS battery is dead or dying.
Won't hurt to change that and run the latest BIOS.

 

[TheoneandonlyMrK]

"The only "problem" is that i must have some erroneous configuration that makes that when activating the DOCP mode of the memory for the 3200Mhz it does a strange thing. After a power loss (unplugged and plugged again), when booting, the fans turn on for 1 or 2 seconds, turn off, do the same again and the third time it starts correctly."

Where are reboots mentioned, when activating DOCP it retrains.

Where are unexpected power losses mentioned.

 

[Bill_Bright]

Where are unexpected power losses mentioned.

You are right - the OP does not use the word, "unexpected". But he/she did say,

After a power loss (unplugged and plugged again)

In ATX PCs, the "normal" method of shutting down a computer is to go through the Windows Start button and selecting Shutdown. It should be noted this actually puts the computer into a sleep/standby state as the ATX standard requires all ATX PSUs maintain a +5Vsb standby voltage in this state. And that +5Vsb voltage is used to maintain +5V on many points on the motherboard, including the CMOS, RAM and other points.

Killing the power via the master power switch or pulling the plug, is, in effect (from the computer's viewpoint) unexpected - even if already "gracefully" shut down.

BUT - even in those scenarios, the CMOS battery is there specifically to hold user settings (changes to the BIOS Setup) and the RTC (real-time clock). If the CMOS battery fails at that job, EVERY TIME a computer completely losses power, user settings like boot order, timings, and even the date and time would be reset back to factory settings.

If this were normal, wouldn't it make sense 10s of 1000s of AMD users would be complaining about this? I note many users around the world use power strips and (like many in the UK) switched wall outlets and routinely kill total power when they are done using their computers for the day.

 

[TheoneandonlyMrK]

You are right - the OP does not use the word, "unexpected". But he/she did say,

In ATX PCs, the "normal" method of shutting down a computer is to go through the Windows Start button and selecting Shutdown. It should be noted this actually puts the computer into a sleep/standby state as the ATX standard requires all ATX PSUs maintain a +5Vsb standby voltage in this state. And that +5Vsb voltage is used to maintain +5V on many points on the motherboard, including the CMOS, RAM and other points.

Killing the power via the master power switch or pulling the plug, is, in effect (from the computer's viewpoint) unexpected - even if already "gracefully" shut down.

BUT - even in those scenarios, the CMOS battery is there specifically to hold user settings (changes to the BIOS Setup) and the RTC (real-time clock). If the CMOS battery fails at that job, EVERY TIME a computer completely losses power, user settings like boot order, timings, and even the date and time would be reset back to factory settings.

If this were normal, wouldn't it make sense 10s of 1000s of AMD users would be complaining about this? I note many users around the world use power strips and (like many in the UK) switched wall outlets and routinely kill total power when they are done using their computers for the day.

He or she said what was said, I'm not trying to hinder him or you.

I agree that once it's trained it's timings it should hold them in its bios but he mentioned activating DOCP.

He didn't say after I've set DOCP every unplug leads to retraining.

I advised the fault he listed not every fault he could see.

I'll leave you to it, I said my piece I've nothing to add.

 

[Bill_Bright]

He didn't say after I've set DOCP every unplug leads to retraining.

You are right. I assumed this was a "continuing issue" and not a "once and done" scenario. If it was "once and done", then I agree that is normal. But if it booted fine after that, then not sure I would call that a "boot problem" or request help on a tech support forum. But then that may be me assuming again.

Perhaps we need more clarification from the OP of the "problem" - assuming there is still one.

 

[Ka0s]

This is NOT normal!

When you first install the RAM (or make changes to the RAM configuration) "training" may be required on the first boot after the change is made. And the training process could take several minute and require yet another reboot. But once complete, it "should" boot normally after that (until another change).

If you have to go through the process of training over and over again, or the day after day boot process requires 3 times to finally boot into the OS, then something is not right! It could be incompatible RAM; is yours listed on the QVL here? Or it could be faulty RAM - have you tried just one stick at a time? Or it could be an unstable setting. Or it could be your CMOS settings are not holding properly.

You say this happens after a full power loss. What happens if you "gracefully" shutdown Windows and your computer normally? Does it boot in one attempt after that?

What happens if you shutdown gracefully, then flip the PSU's master power switch to "0" or "Off" (or unplug from the wall) for a minute or two, then restore power and boot?

Have you reset your BIOS and started over? If me, since they are inexpensive, I would replace the CMOS battery. While they should last many years, some don't. I have even seen DOA, brand new CR2032 batteries right out of the package.

Hi again, sorry about wait time until i could reply.

Yes, its listed on the official QVL > F4-3200C14D-16GTZ.
It is impossible that it is a RAM failure, because if i deactivate the DOCP mode the problem disappears (running bios at default settings). And worked well with my old 1600x.
The BIOS battery is also discarded, i changed it about 3 months ago.

Maybe i didn't explain well, my english is very bad haha:

I always shutdown my pc in the normal way, start > shutdown system.
But the issue only appears when i fully disconnect from AC my pc (power strip switch / PSU's switch / etc..). If i just shutdown and turn on after a few minutes works well.

-

EDIT

For example, when i go to bed i turn off the power strip switch. The next day when i turn it on again the first time, it happens.
And the next day the same, and the next.
Or if I leave the house and turn off the power strip and come back and turn it on, it also happens. I repeat, only when I remove the AC from the pc.

 

[The King]

But the issue only appears when i fully disconnect from AC my pc (power strip switch / PSU's switch / etc..). If i just shutdown and turn on after a few minutes works well.

Then your CMOS battery is dead. Replace it with a new one.

 

[TheoneandonlyMrK]

Hi again, sorry about wait time until i could reply.

Yes, its listed on the official QVL > F4-3200C14D-16GTZ.
It is impossible that it is a RAM failure, because if i deactivate the DOCP mode the problem disappears (running bios at default settings). And worked well with my old 1600x.
The BIOS battery is also discarded, i changed it about 3 months ago.

Maybe i didn't explain well, my english is very bad haha:

I always shutdown my pc in the normal way, start > shutdown system.
But the issue only appears when i fully disconnect from AC my pc (power strip switch / PSU's switch / etc..). If i just shutdown and turn on after a few minutes works well.

-

EDIT

For example, when i go to bed i turn off the power strip switch. The next day when i turn it on again the first time, it happens.
And the next day the same, and the next.
Or if I leave the house and turn off the power strip and come back and turn it on, it also happens. I repeat, only when I remove the AC from the pc.

Have you ever updated the bios of your X370 motherboard?.


Oh and someone or I, should have asked you to fill out your system specifications in your user profile.

People here will actually download your motherboard manual and look into your issue and things like that but only if they know what you have, to help I mean.

 

[Deleted member 185158]

Completely normal for AMD and Intel Asus motherboards that feature memory training algorithms to boot cycle at cold post.

It may happen every cold post, but not typically during a restart or wake from sleep.

In some cases, it may take the board 3 attempts to train the memory, because by default the algorithm gets 3 tries. This is not to be confused with "post fail count" which is also defaulted to 3 tries.