have you ever installed lga3647 cpu ? yu have to turn screws 3 & 4(only one turn per stroke) to the END to pull the leaf spring topmost to the cooler(not faster than about 1.2Nm)
X10SRL-F bad RAM slot or defective CPU?
- Thread starter traderjay
- Start date
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
Probably the board. Traces are easy to scratch. You can try cleaning the ram slots with canned air but beyond that its a bad board.
D
Deleted member 28354
Guest
Thanks much. I tried both 2400 and 2133. Is driving me nuts. It can't be the memory as I swap it around and error does not follow the sticks.
I will do a BIOS factory reset right now.
I will do a BIOS factory reset right now.
forgot to remove a misplaced a motherboard standoff ?
bad cooler, to low pressure, dynatron/noctua ?
bent pin in socket ?
but the MB can also be DOA.
D
Deleted member 28354
Guest
SuperMicro X10DRi, 2* E5-2680v4, 4* 64GB Samsung DDR4 64GB LRDIMM 2400 1.2V 4Rx4 M386A8K40BM1-CRC4Q.
systems does not boot and "Failing DIMM: DIMM location (correctable memory component found)" on the screen.
MB is not in case(no shorts).
All contacts at DIMMs and CPUs are cleaned with alcohol.
This CPUs and DIMMs works fine in another MB.
I just bought this used SuperMicro MB and dont know what BIOS version.
What should I check next?
systems does not boot and "Failing DIMM: DIMM location (correctable memory component found)" on the screen.
MB is not in case(no shorts).
All contacts at DIMMs and CPUs are cleaned with alcohol.
This CPUs and DIMMs works fine in another MB.
I just bought this used SuperMicro MB and dont know what BIOS version.
What should I check next?
"DIMMs work fine in another MB"
Have you memtested them? Probably throwing ECC errors.
this are one bits errors but fixed by ECC. can be the slot, worn out or unclean.
try to run the LRDIMM at 2133, if the corr.errors leave it is not a slot issue.
I tried to clean DIMM slots with canned air. But if I install to E1 the system writes error about E1. If modules in A1 then system writes error about A1. If A1 and E1 then error about A1 and E1. etc. And this dont looks like dirty slots.
can't pass this error message to get to the BIOS.
This modules works with full load(about 250GB RAM used) for two weeks. I'll do memtest.
not with this RAM of course.
i guess the BIOS programs the mem.controller latencies too strange.
overseen this, the board should go further on "correctable memory component found"
afaik it stops only at "uncorrectable memory component found"
it stops at this "correctable memory component found" and I was waiting for about 15 minutes.
IMHO it will go further if at least one module is ok.
this MB does not have clear cmos jumper. I disconnected all cables from MB and removed the battery for 15 minutes. now when this error appeared the system write "[clear CMOS]" on the right side of the screen.
i know. you can short the contacts of the battery holder.
what about NVMEM and SMBios ?
login to IPMI WebGUI and check BIOS version there. old BIOS without V4 should not give that error but who knows.
edit:
or get any cheap V3 CPU and 8GB RDIMM, maybe this works with LRDIMM. then update BIOS with Afu tool
Last edited:
cant reach IPMI. MB powered on for about 30 minutes and I watching DHCP on my router and dont see any new devices. green and yellow leds near IPMI LAN connector are blinkin
cheap E5 v3 already ordered. going to buy cheap 4-8GB DDR4 2133 ECC REG.
IPMI: So, you mention above that you just bought it used. Could the IPMI have a static IP set that is in a range outside your network? The last few boards I got (all from the same source) had static IPs in in the 0.0.0...., which didn't initially show up in my 192.168...& 10.10.10... setups.
Of course, resetting that is gonna be hard if you can't boot enough to set a new IP via the BIOS.
As for booting, sometimes with those big(er) sockets, reseating the CPU helps to fix memory issues. You can have a system that boots most of the way, but then throws a memory error due to a mis-seated CPU (ask me how I know
). 2011-3 isn't as error prone that way as LGA 3647 & other gigantic sockets, but it happens. If the board came to you without a CPU (and you haven't already) maybe also examine the socket really closely with a light for bits of TIM tucked under pins? It probably isn't conductive, but if its a hard bit a pin could deflect under pressure and short against its neighbor. Then when you pull the CPU it springs back and looks totally normal. Not that that has ever happened to me and lost me a half day of testing or anything...