Question I installed an AIO, now I cannot boot with DOCP turned on.

[CBunn]

Hello,

I just upgraded from the stock cooler provided with Ryzen 7 2700x to the H115i Platinum. After doing this I went into bios and set CPU fan to ignore and enable the AIO pump. Saved changes and reset then my system posted into safe mode. I have cleared CMOS and reset all defaults and tried re-enabling DOCP with no success.... I just had to RMA my RAM about 2 months ago so I doubt that could be the point of failure.

Here are the specs of my PC:
Asus Strix X-470 F-Gaming
Ryzen 7 2700x
Corsair H115i Platinum
G Skill (16gbx2) 3200mhz F4-3200C16D-32GTZRX
MSI Armor GTX 1080
Corsair RM850x PSU

If you have any ideas please let me know, It would be greatly appreciated.

 

[Karadjgne]

Put the pump on cpu_fan header, drop the ignore. Nothing in the bios is component specific when it comes to coolers, there's exactly no difference between a stock and the aio, the cpu_fan header reads rpm and nothing else. So unless there's been some other critical change, just swapping coolers out and leaving bios the way it was won't affect anything.

 

[Crosslhs82x2]

Hi we have the same mobo i looked at mine while i was doing a few thing inside the case.
My h110i single wire on the 3pin fan plug is plugged into the cpu header as Karadjgne said and yes cpu fan in monitoring is set to ignore.

The thumb nuts should only be finger tight
If you used a screw driver to tighten,
it could be forcing the cpu into the mobo.

When i put my son' fx-8350 on his mobo and installed the pump it was also finger tight but it would not boot till i backed off the thumb nuts a little bit.

Keep us posted

 

[Karadjgne]

I've never used the 'ignore' in anything, and I've been building for almost 40 years. It's a security feature that trips the cpu into shutdown if the fan reads 0 rpm. With small/stock coolers, that's kinda helpful, with large aircoolers you'll normally have a few seconds before overheat if the fan fails. With aios you'd get less than a minute or so. And mid-game you'd not notice anyways. So it won't hurt to have it on and if the aio pump fails, well it really won't hurt to have it on. It's not something that'll conflict with anything unless you have an issue to start with.

 

[Karadjgne]

DOCP or Direct Over Clock Profile, is ASUS technology that aims to replicate Intel's eXtreme Memory Profile (XMP) feature found on all motherboards running Intel CPUs (assuming the motherboard is not way to old, or is an Manufacture system which they tend to lock the user from changing this option, among other options) .

What it does is get the RAM specs that it gets from requesting it (profile is stored on the memory chip), and then tries to change your motherboard settings memory settings to match the specifications of the RAM so that you can get what you paid for, else it goes with the minimum, safe speed, and it is up to you to configure your memory frequencies and clocks to get the speed mentioned on the RAM stick.

Meaning, if it has to alter voltages, cpu clock speeds etc, it will do so. Having some things enabled/disabled may conflict with that, and you get boot issues.

 

[Darkbreeze]

And, you should NEVER set ANY fan header to ignore, IF there is a device connected to it. Knowing whether or not a fan has stopped functioning is a necessary protection because you WON'T be paying any attention to whether or not there is actually coolant flowing through your AIO and therefore, in spec temps, at all times.

 

[CBunn]

Not exactly sure how it happened, I left the system unplugged over night after giving up on trying to get it to boot with DOCP profile loaded. After work the next day, I got back to it. Plugged in the computer, pushed the power button, the fans kicked on, RGB starts glowing, but there was no action going on, on my display.... No error code, no beeps, nothing.... Im furious, thinking what in the world could I have done to cook my GPU!? So I grab a brand new 1660 TI that I had in the closet, take the clear plastic wrap off the box, unboxed and installed the brand new GPU. Power on, still no freaking display signal. Mmk, great, my 1080 isn't dead, good to know. But I opened this brand new card...

So I have now narrowed it down to 3 components, Motherboard, RAM, CPU....

• Motherboard: Physically the worst PC component you could have to remove in my opinion, but the cheapest of the three.
• RAM: Brand new, I don't want to have to RMA another pair of sticks...
• CPU: I am really considering upgrading to R9, but right now I dont want to fork up the cash for that.

I began the good ole RAM removal troubleshooting. I took the the stick out of A1 and powered on, lights, fans, no action. Okkkkk, possibly bad RAM. I re-insert the RAM into A1 and remove the stick from B1. Powered on.... Voila, we have a post. Holy Toledo Batman, I at least got it to boot, I have at least 1 good stick of RAM and my CPU works.

I boot into windows thinking ok, I can run off 16gb RAM for now. I'll just RMA it and everything will be back to normal in a couple of weeks. After doing stability checks everything is running somewhat ok... I was getting lower scores on synthetic benchmarks but I just figured that it was because I wasn't running dual channel RAM and only on 2133mhz instead of 3200mhz(which is what my RAM is rated for, and has always ran stable).

I was pretty mad about having a dead component and about an hour later I get the idea to try booting with sticks in A1 & A2. Wow, it booted, so that must mean my motherboard has two bad slots(or 1 bad channel).My B1 and B2 DIMM slots are DEAD.... Which really sucks because:

1. This a great motherboard., and i'm used to navigating its bios.
2. I really don't want to have to remove all of my components and basically take my whole system apart.
3. I Don't want to wait for the RMA to process on the motherboard and have no desktop in the mean time.

So what I am going to do is upgrade to an X570 motherboard. Not sure which one yet, but I will figure that out after I get done with sharing my experience from that past few days. I will then most likely buy the 3950X when it comes out. I will use my 2700x for a SFF build that I have been wanting to do for quite some time. I am also heavily considering buying the RTX2080 Super whenever it hits the shelf. So pretty much all new hardware coming within the next few months just because of all this 😁

I am really not sure how I ended up with a defective motherboard. All I was doing was switching coolers. I work in IT and deal with computers all day, every day, but it really hits different when its your own PC that goes down. I was very careful and thorough throughout the whole process. Hopefully it really is my motherboard and I don't have a bad segment on my CPU... Who knows, we shall see once the new mobo comes in.

Thanks for taking the time to assist me, further comments, questions, or suggestions are welcomed!

 

[Darkbreeze]

Why would you have a memory module installed in the A1 slot to begin with?

You have only two sticks from what I can see and that means they should be installed in the A2 and B2 slots, which are the second and fourth slots over from the CPU socket. That goes for ALL dual channel motherboards for as far back as I can remember. Possible for ALL dual channel motherboards, ever.

 

[DMAN999]

I would double check that the cooler is Not too tight as Crosslhs82x2 said that can cause the RAM issues you described.
Also I'd put each stick into slot A2 and run Memtest86 and make sure it is good then repeat that process with the other stick.
If they are both good and the cooler is not overtight I'd put the RAM in slots A2 and B2 as recommended in your manual and try to boot at Optimized Default settings.
If that works then I'd try upping the RAM speed a little at a time until it causes issues again then go back to the highest speed that works.

 

[CBunn]

Why would you have a memory module installed in the A1 slot to begin with?

You have only two sticks from what I can see and that means they should be installed in the A2 and B2 slots, which are the second and fourth slots over from the CPU socket. That goes for ALL dual channel motherboards for as far back as I can remember. Possible for ALL dual channel motherboards, ever.

I had them in A2 and B2 before I started having issues. I moved them to the A1 and B1 when I was having issues booting with DOCP turned on. I did that so I could get a new system configuration post code.

When you have 2 sticks you put them in A2 and B2 to keep them further from the CPU so that they don't get to hot... its only a preventative measure to protect you hardware and does not matter if you have them in the 1 or 2 slots. The 1 slots are just more of a risk of burning your stuff up.

 

[CBunn]

I would double check that the cooler is Not too tight as Crosslhs82x2 said that can cause the RAM issues you described.
Also I'd put each stick into slot A2 and run Memtest86 and make sure it is good then repeat that process with the other stick.
If they are both good and the cooler is not overtight I'd put the RAM in slots A2 and B2 as recommended in your manual and try to boot at Optimized Default settings.
If that works then I'd try upping the RAM speed a little at a time until it causes issues again then go back to the highest speed that works.

I put the stock cooler on and got the same results so Im pretty sure tension isn't the issue. Unless when I initially put it on I mounted it way too tight and physically damaged the motherboard.

 

[Karadjgne]

It's not the heat, the reason A2 and B2 is chosen is because of the memory controller. What you see as A1-A2/B1-B2 from the memory controller point of view is Dimm3-Dimm1/Dimm4-Dimm2. So while you see it as A2/B2 for dual channel, the memory controller sees it as Dimm1+2, primary slots. A1/B1 would be in the secondary position, so there's a miniscule performance loss as the mc calls for data from the primary slots first. Since you are dealing with stuff rated in nanoseconds, what we'd see as miniscule is quite large on that scale.

 

[Darkbreeze]

I was about to say, that has NOTHING WHATSOEVER to do with why you use those slots. And, that will be recorded as a milestone because it's the first time I've ever heard that particular belief. LOL.

You use those slots because that is how the specification is written, and therefore that is how the manufacturers design the boards in order to comply with the standards for the architecture. Has nothing whatsoever to do with heat.

If you look in the motherboard manual for ANY dual channel motherboard, they will all show the 2nd and 4th slots in the population rules as the preferred population guideline. On some boards they will name them differently, calling them DDR4_1 and DDR4_2, for example, instead of A2 and B2, but the population guideline and location is the same regardless. The reason for the use of those specific slots, as Karadjgne has said, has everything to do with how the architecture is designed and nothing to do with heat. If that was the case there wouldn't be any sense in even including or using four DIMM slots, much less six or eight like some HEDT systems have.