ASUS Z10PA-D8 - Dual Socket 2011-3

[Rand__]

No b0 and b7 are not related to memory slots I think - on the other hand what do I know

You don't have any other ram or different cpus to test ? On the other hand you said the board worked fine with these exact modules/cpu's and same bios version before it suddenly stopped working, so it shouldn't be a compatibility issue ...

I am sure you tried a cmos reset, did you?

 

[gb00s]

No b0 and b7 are not related to memory slots I think - on the other hand what do I know

I just thought ...

You don't have any other ram or different cpus to test ? On the other hand you said the board worked fine with these exact modules/cpu's and same bios version before it suddenly stopped working, so it shouldn't be a compatibility issue ...

I only have the 2630 v3's and a 2673 v3. Both the same issue. I tested it already. Correct, the last two boards worked and just went off at some point. But I have one board which is totally fine and doesn't have/had any hiccups which I did not buy from the eBay seller. I updated the bios there as well but not(!!!) the firmware. All other boards I bought from the eBay-seller I updated bios and firmware. All dying. Hmmm, .... Don't get me wrong I suppose it's not(!!!) the guy's fault. Maybe there's a fw issue.

I am sure you tried a cmos reset, did you?

Of course, I did.

I know some other guys here bought the same board from the same seller on eBay. Did they also upgrade the firmware? Would be interesting.

 

[Rand__]

@RageBone - do you have any further ideas?

 

[gb00s]

After reloading some firmware and jumping back to the current firmware, I get readings from the iKVM again. But unfortunately, no progress on the boot as it still stops at 60 or b0. Of course readings for cpu2 are not available as only 1x cpu is installed.

 

[alex_stief]

Let's hope that Vcore1 is a wrong sensor reading. Almost 1.8V would kill modern CPUs more or less instantly.

 

[JanCerny]

@gb00s Maybe it is shorted by dust. Did you try to clean whole board?

 

[gb00s]

Around 12.30am tonight, I got the board booting populated with one and shortly after two sticks on slot A1 and C1 if I remember correctly / Link: https://drive.google.com/file/d/1YG1C-MM1IxQLYvYWFsWw01BLkoF7RgG5/view?usp=drivesdk .

It ran for around an hour until I powered it off and populated all 4 slots of socket 1. Nothing. Same problems. So I went to the company tonight, took out the ram from a working Z10PA-D8, and set the ram into all 4 slots. several power cycles till postcode 04, then 19, then b0 and b7/b9, CMOS resets and .... Working again ..... Till this morning. The board just shut off itself again. I did not touch it overnight. It was just running. Different PSU this time, SM 920P-1R and different (authorized) ram.

Now it's just as dead as yesterday around midday. No postcode. Blank screen.

EDIT: In regards to the comments with suspected shortages.

 

[Rand__]

so you have basically replaced everything and yet its showing this flaky behaviour... quite mysterious

should'nt be temp related should it? ie its not working when its warmer?
Electrical charge buildup somehow?
Anything in the log before it powers down? or just gone?

 

[Jernox]

I also got the same board from the same seller, and it took me quite a while to get it to work. I guess in hindsight it could be caused
by memory incompatibility, although now I have memory modules with exactly the same chips, but those modules are listed on the QVL, whereas my first modules weren't. Tried to contact the seller several times, but I never got a reply.

I haven't installed an OS/hypervisor yet, but have booted into some live environments to check some hardware diagnostic tools. Before I'm going to install, I would like to run some stress tests first, but my eye caught a very high vcore1 and vcore2 voltage. In the BIOS and HWiNFO both vcore voltages are around 1.8v. Although it seems highly unlikely to me that the readings are correct, I still want to be sure about the actual voltages before I start any form of stress testing.

I saw gb00s' screenshot above which showed a similar reading, and I was wondering if there are more people here with such high readings.

My CPU temperatures are just fine btw (2 x 2630l v3).

 

[JanCerny]

In the BIOS and HWiNFO both vcore voltages are around 1.8v. Although it seems highly unlikely to me that the readings are correct, I still want to be sure about the actual voltages before I start any form of stress testing.

Have you voltmeter?

 

[Jernox]

Have you voltmeter?

I do, but haven't looked into measuring vcore with it. For now I prefer a software based solution.

 

[JanCerny]

Software solution is inaccurate. I recommend measure it on inductor wire.

 

[Jernox]

Software solution is inaccurate. I recommend measure it on inductor wire.
View attachment 14637

Thanks, although it seems that my cooler is really preventing easy access with the probes.

I have to see where else I can measure, but if there's too much of a risk of shorting two points, I'd rather stay away from that method .

I also don't need a super accurate reading but as of now, the measurements are way off, thus useless.

 

[gb00s]

Just started working on it again. I just want to make something clear. I'm not talking against the seller here. I had always good communication and he immediately sent out and exchanged new boards after I claimed something is not working. I bought 4x boards from him and 1x from another eBay seller. All 4 from him were exchanged once. Yes, it's bad to have all 4 exchanged, but at least I had no problems getting them exchanged.

Should'nt be temp related should it? ie its not working when its warmer?

I remember there where 4 entries of temp warnings in the log before I upgraded bios & firmware. Facepalm ... I should have saved these log entries. I did not think about it as all boards had the entries when they were delivered. Some had fan errors. The 2x last boards had Noctuas on it, with proper Noctual thermal paste. I never saw readings > 60C even with Pime95 tests. Otherwise, the boards got enough airflow. The one in front of me is not hot anywhere.

Electrical charge buildup somehow?

I thought about it but I can not see from where.

Anything in the log before it powers down? or just gone?

Logs are always empty. Audit just filled with entries from my access through iKVM.

EDIT: I have the board back up again with to the 00 postcode and the boot entries on the screen is telling me 'Initializing BMC' and BMC is ready' ... Then powers off. I can restart it and stays up. The weird thing is the postcode then shows a loop between '00 as successful boot and '04' ... If I load 'PartedMagic' as a live-system, it shows postcode 'FF'. It's not described anywhere in the postcode overview of this board. What does postcode 'FF" mean? Temps are quite normal. But I still can't read any FPU data which was the case when I got the board.

 

[Rand__]

If you have a working on in the office, can you compare iKVM values? Especially voltage since that seems to be high (or quite inaccurately measured).
If you have now replaced all 4 boards - did the other ones work fine after replacement ? All replaced for the same reason (stop working)? You sure you've not gotten one that you sent back?

 

[gb00s]

If you have a working on in the office, can you compare iKVM values? Especially voltage since that seems to be high (or quite inaccurately measured)

I will check tomorrow. I'm not stressing security guard again tonight But I have the feeling we may confuse ourselves with VDDQ_AB_CPU and Vcore. I saw some other guys posting Vcore for these series of cpu's ~1.8v.

If you have now replaced all 4 boards - did the other ones work fine after replacement ?

Yes, ...

All replaced for the same reason (stop working)?

The last two have/had the same issue.

You sure you've not gotten one that you sent back?

Yes, because whenever I claimed a board is not working, he immediately (almost the same date) sent me another one.

EDIT1: SO the board runs again for 15min or so, but there's this new postcode I never saw before on this board and which is not documented >> 'FF' ... Hmmm .... But do you want such a flaky board running in a project? I may go with a WTB: Supermicro X10... To much trouble with these.

EDIT2: Seems like from other forums FF on an ASUS means FAULT FOUND. Ram is in QVL list. Processor is ok. Cpu fan is running. Hmmm ...

 

[Rand__]

Ok, but that then tells us that
a.) he is used to boards not working (or would have raised some discussion after the second/third replacement)
b.) failure is happening quite often (at least under some specific conditions we have not found yet) (since you have 4 boards with issues already)

Have you asked the seller if he has any idea re possible root cause?

 

[gb00s]

Have you asked the seller if he has any idea re possible root cause?

I would have to re-read the whole conversation. But there was never a real discussion from their side. You could argue, it comes from a render farm and was 4yrs under their 'load'. But aren't these machines normally used for gpu workload?

EDIT: Time ... 11.33 pm ... Board shuts itself off again. Last state when the board died again. Logs empty.

 

[Rand__]

well discussion with the seeller only to get further insights...
and if only to see how many replacements he is willing to send out till they all work ..

 

[Jernox]

I wonder what I did wrong when contacting the seller. I explained the situation and just asked for some advice, and I didn't rush at all, there were 8 days in between my first and second message.

I had the intention to buy several of these mainboards, but due to my experience at that moment with the seller and troubleshooting, I was't really tempted to order some more.