The Intel Atom C2000 Series Bug – Why it is so quiet

[Patrick]

I wonder if there's a way to tie failing nodes by week of manufacturing so we could find which weeks to avoid. But it seems like it's unlikely to fail for us

I suspect Intel is doing this.

Also - the silicon re-spin parts are not out yet which means that is not going to help.

 

[voodooFX]

I'we got my a1sri 2558f on January 10th
Is there a way to check if it has the "platform workaround fix" already applied?

Anyway this bug is something that just kills the board while it's running or what?

 

[Patrick]

I'we got my a1sri 2558f on January 10th
Is there a way to check if it has the "platform workaround fix" already applied?

Anyway this bug is something that just kills the board while it's running or what?

I do know a common symptom is that the board will not post and you can get diagnostic information to see if it is indeed the issue via console.

Let me see if I can get a script.

 

[voodooFX]

Thanks!

So the fix is never reboot, easy

 

[Patrick]

Thanks!

So the fix is never reboot, easy

I am not sure that is accurate, just that the failure behavior is a failure to post.

 

[dealcorn]

Has Supermicro provided any guidance beyond the standard RMA procedure for possible warranty issues? Is an extended warranty possible or likely for affected motherboards? Is it possible to speculate whether the bug is more susceptible to 24/7 operations vs normally off operation with daily, or more frequent, WoL?

 

[shlagevuk]

I've bought an Asrock 2750d4i in 2015, it was dead 16 month latter, with symptom indicating a CPU failure (ipmi was still available).

I hope there will be some kind of official communication about this so I can return my board for warrantly against hidden defect (French law)

 

[Patrick]

@shlagevuk at 16 months it is very unlikely to be this issue.

 

[fta]

Anyway this bug is something that just kills the board while it's running or what?

That is my experience. My NAS was hung when I got up this morning. Power cycled it, and it won't boot now. It lasted 2 years and 1 week.

 

[Yamabushi]

That is my experience. My NAS was hung when I got up this morning. Power cycled it, and it won't boot now. It lasted 2 years and 1 week.

Uhhh... that is definitely not what I wanted to hear. I was hoping I'd only be at risk on reboots.

 

[Diavuno]

Has Supermicro provided any guidance beyond the standard RMA procedure for possible warranty issues? Is an extended warranty possible or likely for affected motherboards? Is it possible to speculate whether the bug is more susceptible to 24/7 operations vs normally off operation with daily, or more frequent, WoL?

Nothing I've seen outside of RMA

They are pretty awesome with RMA stuff, just give them call.

 

[KaHaR]

Nothing I've seen outside of RMA

They are pretty awesome with RMA stuff, just give them call.

I sent an RMA request via their support page referencing the Intel errata and they sent me an RMA# the next day. This was for a board I got off of eBay too.

 

[Diavuno]

I sent an RMA request via their support page referencing the Intel errata and they sent me an RMA# the next day. This was for a board I got off of eBay too.

NICE

 

[Kev]

This has been my RMA experience with super micro as well. RMA accepted and cross shipped me a board in 2 days and now I will bring back my old board and drop it off instead of shipping it since I am local. Cool!

 

[Evan]

I asked some very specific questions to 2 different vendors and they said they either don't want to or can't answer my questions. Which is rather frustrating as just for interest I would love to know how the fix is managed.

 

[vvu]

Proactively RMA'd a board through supermicro. I haven't had a chance to install it yet but I think I found Intel's platform fix. Its very minor, two bodge wires.

Sorry for bad photo, I just don't have patience to setup a rig to photograph PCB's.

 

[vvu]

Just an update on this. I swapped the board today. Unless the board revs are different I can confirm the bodge wires did not exist on my old board.

 

[Yamabushi]

Just an update on this. I swapped the board today. Unless the board revs are different I can confirm the bodge wires did not exist on my old board.

Very interesting!

 

[fl0w]

Hi, sorry for digging up this topic, but do you think Supermicro is still offering mobo replacement? I just sent them a RMA request, I hope they will offer a mobo swap.

 

[britinpdx]

do you think Supermicro is still offering mobo replacement? I just sent them a RMA request, I hope they will offer a mobo swap.

I sent SM (USA) an A1SRi-2758F motherboard for repair on Feb 13th. I did not request an advanced exchange, just sent in the board for "repair for AVR54 bug". SM shipped back the board on Feb 22nd.