You got it right at the first time, the myth of 1+ g gold per ppro is
kept alive by eBay sellers that sells to less informed buyers.
PPros is closer to 1/3 of a gram Au per CPU (might vary up and down a
bit depending on which CPU / cache combo. Most of the gold is in the
bond wires and the solder holding the chips on place (a silicon-gold
braze ).
http://goldrefiningforum.com/phpBB3/viewtopic.php?f=60&t=15269&hili…
I'm a gold refiner / scrapper / classic computer collector, all on a
hobby basis, so my workshop is fast filling up with computers that I'm
not allowed to refine.
... there's some mixed feelings there.
:-D
G?ran
William Donzelli skrev den 2014-11-25 07:18:
Oops, I am a bit too tired, and something was bugging
me about your
calculations.
PPros have more than a third of a gram - they are closer to a gram per unit.
--
Will
On Tue, Nov 25, 2014 at 1:08 AM, William Donzelli <wdonzelli at gmail.com> wrote:
>> Let's say that a Pentium Pro has a third of a gram; a troy ounce is
>> 31.1 grams; so 94 Pentium Pro to the troy ounce; 4.5 pounds is about
>> 65 troy ounces; so 6,182 Pentium Pro might have that much gold.
> There are 12 troy ounces to the pound, not 16.
>
> Anyway, yes, you are talking about something like 4800 PPros
> equivalent. Now compare the plated surface areas of a PPro compared to
> a number of the pins on the backplane - you will find that a gold
> plate area on a PPro (mind you, there is a fair amount under the
> cover, too) equates to a surprisingly small amount of old IBM
> backplane pins. Remember, the pins are long, and plated on four sides
> - and probably had a thicker plate that on PPros.
>
> And yes, the 75 was a really huge machine. Gold on the backplanes,
> gold on the connectors (early Bus&Tags have a very heavy plate), gold
> in transistors. I would not be surprised if there was gold in the lamp
> sockets.
>
> --
> Will