[Tfug] Interchangeability of VRM's?

[Bexley Hall]

Hi Zack,

On 12/10/2012 12:58 PM, Zack Williams wrote:
> On Mon, Dec 10, 2012 at 12:52 PM, Bexley Hall<bexley401 at yahoo.com>  wrote:
>> Similarly, does the VRM connect to the host/motherboard by
>> some standardized interface?  Or, is the VRM seen as an extension
>> of the motherboard?
>
> How old of hardware are we talking about?

I think about 6-7 years.  Dual 3.4G Xeon's.

> The last unit I had that needed a separate VRM was my first x86 box
> (1997, dual Pentium Pro... those were the days).

My first "PC" was a 386.  Before that, a Z80 (CP/M) -- though with
0.5M of RAM (astonishing, back then!)

> My recollection is that they're processor specific, but standardized
> across logic boards that support those processors.

So, changing clock speed would (probably) utilize the same VRM.
And, a Dell machine supporting a Frazzle9000 CPU would use the
same *actual* VRM (same pinout, specifications, etc. -- though
possibly made by a different garage shop in CHina...) as an
Acer machine using the same Frazzle9000 (?)

More to the point:  if I find a Frazzle9000, I can pull *it*
and it's VRM and add them as a *second* processor in another
machine (perhaps made by a different manufacturer) that has a
*single* Frazzle9000 in it, currently (?)

Or, if I found a Frazzle9000 operating at a higher clock speed,
I could replace the Frazzle9000 in my machine with that device
without worrying about the VRM (?)

(sigh)  I am amazed that there is *any* degree of standardization
*within* motherboards!  Some grand conspiracy?  :>

Thanks!
--don