Question SSD issue: Current transfer mode is PCIe 4.0 x2, when it should be PCIe 4.0 x4.

[alexandrus_e]

The normal thing would be to think that the motherboard is not compatible with PCIe 4.0 x4, however it is compatible. Additionally, the processor is also compatible with PCIe 4.0 x4, so I have no idea where the problem lies.

The components I'm using:
SSD: Kingston KC3000 1 TB
Motherboard: B550M DS3H AC (rev. 1.0/1.1/1.2/1.3)
Processor: AMD Ryzen 5600X
Graphic card: RX 6600

Update: The M.2 is placed in the M2A_CPU slot, so I think the problem may be on the motherboard (factory defect) but it would be very rare.

 

[rgd1101]

latest motherboard bios?

 

[alexandrus_e]

latest motherboard bios?

F17a.

 

[Bob.B]

The normal thing would be to think that the motherboard is not compatible with PCIe 4.0 x4, however it is compatible. Additionally, the processor is also compatible with PCIe 4.0 x4, so I have no idea where the problem lies.

The components I'm using:
SSD: Kingston KC3000 1 TB
Motherboard: B550M DS3H AC (rev. 1.0/1.1/1.2/1.3)
Processor: AMD Ryzen 5600X
Graphic card: RX 6600

Reseat the m.2.....test.

Look through the bios for a m.2 lane option.

 

[Maxxify]

As Bob said, reseat. A loose x4 PCIe SSD will report as x2 if certain pins are not in contact.

 

[alexandrus_e]

As Bob said, reseat. A loose x4 PCIe SSD will report as x2 if certain pins are not in contact.

Done, it still doesn't work, but thank you for responding. Just in case it helps, I am attaching an image of the back of the SSD. I have noticed that there is a small defect on the pin in the highlighted area in the image, although it is not very clear.

 

[Bob.B]

Done, it still doesn't work, but thank you for responding. Just in case it helps, I am attaching an image of the back of the SSD. I have noticed that there is a small defect on the pin in the highlighted area in the image, although it is not very clear.

You might want to try this ssd in another pc and see what it reports there.

 

[Maxxify]

Done, it still doesn't work, but thank you for responding. Just in case it helps, I am attaching an image of the back of the SSD. I have noticed that there is a small defect on the pin in the highlighted area in the image, although it is not very clear.

The last two pins (offset per side) are used for voltage (3.3V on this one) and then module type (pin 75 for SSD is always 0). You can see information on the drive in UEFI via NVMe Configuration, under Settings and IO Ports. I'm not sure if this is a software issue but that can be tested (bootable Linux, install KDiskMark) or hardware (a NVMe to PCIe adapter would at least show if 4 lanes can work on PCIEX4).

 

[alexandrus_e]

The last two pins (offset per side) are used for voltage (3.3V on this one) and then module type (pin 75 for SSD is always 0). You can see information on the drive in UEFI via NVMe Configuration, under Settings and IO Ports. I'm not sure if this is a software issue but that can be tested (bootable Linux, install KDiskMark) or hardware (a NVMe to PCIe adapter would at least show if 4 lanes can work on PCIEX4).

Update: I bought an M.2 to PCIe adapter and these are the results I got:
PCIE adapter:
-CDInfo:

-BIOS:

M2A_CPU slot:
-CDInfo:

-BIOS:

 

[drtweak]

So all 4 lanes are working normally then. Might want to check to make sure no debris on the slot maybe?

 

[Maxxify]

Good to see the adapter works. Suggests issues with the M.2 slot or CPU - M.2 pins, CPU pins, or traces. If you had another SSD to try, that might help confirm it's not just a compatibility issue with the KC3000. CPUs/platforms can negotiate down with certain power settings but usually this is 4.0 -> 3.0 rather than bifurcating lanes. There are actually issues like this on the Z690 platform (Intel) that's kind of wonky with PCIe 5.0, but that does not apply here obviously.