Z790 Dark Hero - A device which does not exist was specified

[paulella]

Recently I have run into the error - 

A device which does not exist was specified. 

This seems to be happening for 3 of my 5 m.2 NVME devices. At first, I thought it was the device itself but it's hard to believe that 3 different m.2 NVME devices failed simultaneously.  Has anyone else encountered this problem?

Thanks

[Silent_Scone]

Hi @paulella 

Did the issue occur suddenly? Can you state the make and model of drives?

1. Flash to BIOS 1302

2. Retest at Optimised Defaults (do not enable XMP Overclocking).

3. Remount the CPU to ensure pin-pad contact and inspect the socket and pin-pad for any damage or debris. Remount cooler whilst ensuring for consistent mounting pressure.

[paulella]

These are the drives 

Storage
Samsung SSD 990 EVO 2TB 
Samsung SSD 970 EVO Plus 2TB 
Samsung SSD 970 EVO 1TB 
Samsung SSD 990 EVO 2TB 
Samsung SSD 970 EVO Plus 1TB 

The 3 drives fail randomly, in different combinations, but always the same 3. They can be working one moment then fail the next. 

1. Bios already at 1302

2. I don't have any overclocking activated. 

3. I haven't had any issues with the CPU, it has passed all the P95 -CPU Tests, and a visual inspection doesn't indicate any damage or debris. 

The problem has been intermittent since I got the motherboard in March, now it is happening every time I reboot the PC. 

Thanks! 

[Silent_Scone]

1. I would advise against running P95 due to the amount of current pulled. Moreover, Prime95 only proves the system is stable at running Prime.

2. Please state which slots the afflicted drives are connected to.

3. If using a CPU contact frame, remove and retest with the stock retention mechanism. An intermittent fault could still indicate improper pin pad contact.

4. You can try the beta BIOS listed here which contains a more recent microcode (0x125) from Intel.

5. As several drives are connected, I would limit to those essential and see if the behaviour persists. Please use Samsung Magician to inspect the health of the afflicted drives.

[paulella]

I use water cooling for the CPU and GPU on separate loops. I have detatched the CPU loop, removed the cooling block from the CPU and reseated the CPU as requested. Nothing has changed when I put the PC back together and rebooted. 

I am running Samsung Magician (latest version 8.1). It shows the drives that are available (those not currently having the error) and it shows they are all heathy with no issues. If I reboot and those affected drives are available this time around, it shows them as healthy as well. All drives have the latest firmware updates according to SM. 

Because the drives might not be in the correct order that their drive letters indicate I will need to pull them out to determine which slots are problematic. I know that Drives C and F are never affected. Only D, E and G. I just don't know what slot each drive equates too. Although I have a feeling they are M.2_3,  M.2_4,  and M.2_5.  

I plan to take them all out apart from C and then install them into M.2_2 one at a time to check they are operational first then determine if my assumption is correct about the slots. Stay tuned!