mike wrote:
On 12/7/2018 7:28 PM, wrote:
On Friday, December 7, 2018 at 6:01:08 PM UTC-8,
wrote:
Hi,

I've acquired 2x4GB new memory from Crucial as I found it to be
compatible with my PC(ASUS CM6850) on crucial.com.

It suggested `DDR3L-1600 UDIMM` and that is what I got (results
slightly changed but still align:
https://www.crucial.com/usa/en/compa...#MEMORYFilters)


Now, I popped them into the two empty banks in my PC, turned it on,
the fans spun up but both screens stayed black, I removed one bank,
repeated the power up procedure but got the same result. I then
removed both new banks and now it powered up again properly as it's
supposed to.

Does this mean the memory I got is for some reason incompatible with
my system?
Why?

Here's some images:
The label on my PC: http://gdurl.com/qN7c
The existing RAM on the main board: http://gdurl.com/vHcm
My new memory that I wanted to add: http://gdurl.com/rOee

I think your problem is that you have PC3 and PC3L memory.
The Samsung is likely 1.5V, the Crucial is 1.35V
As I understand it, the low voltage memory is supposed to be
backward compatible, but isn't always. Never could tell whether
the PC3L was 1.5V tolerant or whether the motherboard could tell
and reduce the voltage as required.

I have the same problem with a laptop.
My PC3L is labeled for Apple computers, so that may be a different/another
issue as well.

Server motherboards could definitely run that stuff
at the stated 1.35V. That's why that memory was invented,
to save power on servers.

There are also a few desktop motherboards, with
programmable VDimm for that purpose.

But some older kit, just runs it at 1.5V, and the
memory was supposed to be designed to tolerate both
conditions.

*******

There's two ways they could have policed the voltage.

1) The value stored in the SPD on the DIMM.
They could have defined a way to store the value electronically,
and the BIOS evaluates the SPD contents, before
"turning on" the DIMMs. The memory doesn't start
right away. The BIOS uses some register-only code
as part of the bring-up process. One of the last
things the BIOS does, is do the special CAS value
programming cycle, before the memory starts accepting
transactions.

2) You can use the slots sawed in the DIMM connector,
to prevent a 1.35V stick from going into a
1.5V motherboard. But I don't get the impression
they made any attempt to do it that way. AFAIK
they're both the same edge card pattern. Maybe you
could research that by looking at some pictures
of 1.35V and 1.5V products and compare the connector.

You can dump the contents of the SPD using CPUZ results
file. But decoding the values, deciding whether a voltage
field says 1.35V or 1.5V, that takes a JEDEC document
to do the decoding. And then I have to remember what
my JEDEC password is :-) Or whether I even have one :-)

Paul