MP 1,1-5,1 cMP won't recognize more than 6 DIMM's

[Dwight__SCHRUTE]

I have a 4,1 -> 5,1 cMP. Specs are: 2x x5680, RX580, one NVMe boot + several HDD's and 6 sticks of 1333MHz RAM.

I have 8 sticks in total. Every single stick in isolation is addressed in slot 1. However, when I start to put more sticks of RAM into the machine, the computer may or may not recognize them. To help troubleshoot, I numbered each stick 1 through 8. The computer will not boot if there are more than 6 sticks present. The computer will boot with sticks in slots 2,3,4,6,7,8 but NOT with sticks in slots 1,2,3,5,6,7 but as I mentioned, the computer will boot with a single stick in slot 1 so it appears that the slot is in good working order. Also, I can only get the computer to address a max of 6 DIMMs only if certain DIMMs are in certain slots. For example, if DIMM 1 is in slot 2, the computer will recognize it. If DIMM 1 is in slot 3, the computer won't.

I experienced the same problem before I upgraded the processors (the computer came with two x5550s) so it doesn't seem that the problem is a CPU one.

I was thinking about buying new sticks, but every single stick I have works when used in isolation so the problem isn't with my ram modules. Has anyone ever run into this problem before?

 

[tsialex]

I have a 4,1 -> 5,1 cMP. Specs are: 2x x5680, RX580, one NVMe boot + several HDD's and 6 sticks of 1333MHz RAM.

I have 8 sticks in total. Every single stick in isolation is addressed in slot 1. However, when I start to put more sticks of RAM into the machine, the computer may or may not recognize them. To help troubleshoot, I numbered each stick 1 through 8. The computer will not boot if there are more than 6 sticks present. The computer will boot with sticks in slots 2,3,4,6,7,8 but NOT with sticks in slots 1,2,3,5,6,7 but as I mentioned, the computer will boot with a single stick in slot 1 so it appears that the slot is in good working order. Also, I can only get the computer to address a max of 6 DIMMs only if certain DIMMs are in certain slots. For example, if DIMM 1 is in slot 2, the computer will recognize it. If DIMM 1 is in slot 3, the computer won't.

I experienced the same problem before I upgraded the processors (the computer came with two x5550s) so it doesn't seem that the problem is a CPU one.

I was thinking about buying new sticks, but every single stick I have works when used in isolation so the problem isn't with my ram modules. Has anyone ever run into this problem before?

First thing is to check if all DIMMs are the exactly same type and rank. You need DIMMs of the exactly same type (UDIMMs or RDIMMs) and 2Rx4 rank. You can't mix-match.

Second thing is to check the sockets for any missing or bent pins.

 

[Snow Tiger]

I have a 4,1 -> 5,1 cMP. Specs are: 2x x5680, RX580, one NVMe boot + several HDD's and 6 sticks of 1333MHz RAM.

I have 8 sticks in total. Every single stick in isolation is addressed in slot 1. However, when I start to put more sticks of RAM into the machine, the computer may or may not recognize them. To help troubleshoot, I numbered each stick 1 through 8. The computer will not boot if there are more than 6 sticks present. The computer will boot with sticks in slots 2,3,4,6,7,8 but NOT with sticks in slots 1,2,3,5,6,7 but as I mentioned, the computer will boot with a single stick in slot 1 so it appears that the slot is in good working order. Also, I can only get the computer to address a max of 6 DIMMs only if certain DIMMs are in certain slots. For example, if DIMM 1 is in slot 2, the computer will recognize it. If DIMM 1 is in slot 3, the computer won't.

I experienced the same problem before I upgraded the processors (the computer came with two x5550s) so it doesn't seem that the problem is a CPU one.

I was thinking about buying new sticks, but every single stick I have works when used in isolation so the problem isn't with my ram modules. Has anyone ever run into this problem before?

1 ) Reset NVRAM of your Mac . It's the old hold the PRAM keys continuously at start until your Mac chimes three times before going into an OS session . This might do the trick . If not :

2 ) Please take a picture of all your memory modules with the model labels face up . We need to see this clearly .

3 ) Is your Tray clean ? Dust might have gotten into the slots and prevented a gold tooth on the memory from making contact . Or a small piece of metal got into the slot and is shorting it . This happened to me in the last week with a client's Mac .

4 ) Inform us of any red lights inside your Mac that continuously stay on during an OS session .

 

[Macschrauber]

I bet for bad contacting of lidded Xeons

 

[MacUser2525]

I bet for bad contacting of lidded Xeons

Indeed an extra small turn on the heat sink screws may help it work or a loosening, it did for me when replacing the processors and not all slots worked. What a scummy method of doing it Apple choose for that model.

 

[Snow Tiger]

I bet for bad contacting of lidded Xeons

that is a possibility , as his Xeon has an in-silicon memory controller .
[automerge]1581261088[/automerge]

Indeed an extra small turn on the heat sink screws may help it work or a loosening, it did for me when replacing the processors and not all slots worked. What a scummy method of doing it Apple choose for that model.

the OP would be well advised to grab an AA flashlight and take a look at the four CPU Heatsink fastener positions relative to their corresponding CPU Tray's HS standoffs . There should be no gap between them and once verified , loosen each fastener by a quarter turn to prevent an over-tightening situation . This will prove useful if the Mac is shipped or harshly moved .

 

[Dwight__SCHRUTE]

First thing is to check if all DIMMs are the exactly same type and rank. You need DIMMs of the exactly same type (UDIMMs or RDIMMs) and 2Rx4 rank. You can't mix-match.

Second thing is to check the sockets for any missing or bent pins.

Hi. Thanks for the reply.

All the DIMMs have identical markings on them. I have looked into the slots and there do not appear to be any bent pins.
[automerge]1581273137[/automerge]

1 ) Reset NVRAM of your Mac . It's the old hold the PRAM keys continuously at start until your Mac chimes three times before going into an OS session . This might do the trick . If not :

2 ) Please take a picture of all your memory modules with the model labels face up . We need to see this clearly .

3 ) Is your Tray clean ? Dust might have gotten into the slots and prevented a gold tooth on the memory from making contact . Or a small piece of metal got into the slot and is shorting it . This happened to me in the last week with a client's Mac .

4 ) Inform us of any red lights inside your Mac that continuously stay on during an OS session .

Hello. Thank you for the reply.

1.) I have tried that, though the last time I did was before I swapped CPU's. I'll try it again.

2.) I have once again verified that all DIMMs have the same markings on them, but I will take pictures and post.

3.) The tray is clean. I blew the slots out with canned air and cleaned all the contacts on the DIMMs with rubbing alcohol.

4.) No visible red lights except for the one on my NVMe sled.
[automerge]1581273210[/automerge]

I bet for bad contacting of lidded Xeons

Hello. Thanks for the reply.

I used a pair of delidded Xeon's for the swap. However, I've had this problem since long before I did the processor swap.
[automerge]1581273338[/automerge]

Indeed an extra small turn on the heat sink screws may help it work or a loosening, it did for me when replacing the processors and not all slots worked. What a scummy method of doing it Apple choose for that model.

Hey. Thanks for the response.

When I initially reseated the heatsinks, I had to carefully tighten/loosen screws five or six times before the computer would boot. The heatsinks were never removed previously so I don't think it is an issue, but I may be wrong. I'll try this if no other fixes work. Thanks for the idea though.

 

[Dwight__SCHRUTE]

that is a possibility , as his Xeon has an in-silicon memory controller .
[automerge]1581261088[/automerge]


the OP would be well advised to grab an AA flashlight and take a look at the four CPU Heatsink fastener positions relative to their corresponding CPU Tray's HS standoffs . There should be no gap between them and once verified , loosen each fastener by a quarter turn to prevent an over-tightening situation . This will prove useful if the Mac is shipped or harshly moved .

Frightfully sorry. Would you mind explaining the bit about "HS standoffs" and the gap some more?

 

[Snow Tiger]

Frightfully sorry. Would you mind explaining the bit about "HS standoffs" and the gap some more?

I don't have pictures from the right perspective , but I can show you what the two separate parts look like .

HS is short hand for heatsink , in the industry .

The CPU Heatsinks each have four retained fasteners . They insert into the Tray's standoff holes .


Here are the CPU Tray heatsink standoffs for CPU A of your Nehalem DP Mac Pro . This is for a factory 2010 - 2012 model . A 2009 model would look a little different . The CPU Tray heatsink standoffs are circled in red .

And here is a CPU heatsink from a Nehalem DP Mac Pro . The CPU Tray heatsink standoffs are circled in red . The CPU Heatsink fasteners are circled in green . They drop down when the heatsink is held level . They also lower further when tightened with a driver , once placed onto the CPU socket .

Unfortunately , I don't have a close up picture of the connection of these two components . But observe this connection area while you are tightening the fasteners .