Jonathan Billings wrote:
> I cut out the text of dmidecode that weren't necessary for the point:
>
> On Mon, Sep 08, 2014 at 01:52:43PM -0400, Bob Metelsky wrote:
>> dime910 /home/robert :( # dmidecode -t 17
>> # dmidecode 2.12
>> SMBIOS 2.3 present.
>>
>> Handle 0x1100, DMI type 17, 27 bytes
>> Memory Device
>> Locator: DIMM_1
>> Part Number: 64T256020EU2.5C2
>>
>> Handle 0x1101, DMI type 17, 27 bytes
>> Memory Device
>> Locator: DIMM_3
>> Part Number: CM2X2048-6400C5
>>
>> Handle 0x1102, DMI type 17, 27 bytes
>> Memory Device
>> Locator: DIMM_2
>> Part Number: CM2X2048-6400C5
>>
>> Handle 0x1103, DMI type 17, 27 bytes
>> Memory Device
>> Locator: DIMM_4
>> Part Number: 64T256020EU2.5C2
>
> I couldn't find the Dell Service Manual for a 'Dell 770' but I was
> looking at the Service Manual for the Optiplex 760, and I saw that the
> way that the modules are paired isn't obvious. It looks like you've
> got 2 2G memory modules from 64T256020EU2.5C2 (Kingston?) and 2 2GB
> modules from CM2X2048-6400C5 (Corsair?). I suggest finding your
> service manual, and make sure that the modules are placed in the DIMM
> slots and the same vendor RAM is paired with its partner.
>
> It sounds to me like there might just be something weird going on with
> how the memory is installed. I'm assuming you don't have an
> artificial limit in the kernel command line or anything obvious like
> that.
>
Oh, Ghu.... I don't know about lower-end desktops, but in servers, they
*MUST* be *identical*, even to the point of not being able to mix dual
rank with quad rank, even if everything else is the same.
mark