chipfans: help me solve the MediaGX/China problem

[CPUShack]

I have been tracking production codes on MediaGX cpu's for about 3 years now.

Suddenly we get a bunch from China, and none of my codes make sense any more.

Can you provide me with a pic of as many BGA MediaGX's (Various speeds) as you can to help sort this out.

Were these possibly made in China?

[sammyc]

And if you could also get back to me regarding the CPU's missing from my parcel...

[pinkie]

[chipfans]

[CPUShack]

Here is an id guide I wrote.

http://www.cpushack.net/MediaGX.html

before the chinese CPU's came n all dates and die codes matched (the later the date, the later the die revision), now they dont.

[chipfans]

[jd]

Hi bud, I actually can't help much with these codes ,but I do have a question regarding this: "ANY MediaGX, GX1, GXm or GXLV with a speed over 333 is a fake."(which you mention in your text).

How is it that these CPU's that are considered "fake" be just that , if they were manufactured?...I really am trying to understand that, but when I hold these in my hands, it's hard for me to see why they would be fake.

Now what I do understand is the whole clocking issue where they stopped at ....what is it 300mhz?...around there...

Basicaly I need to be reinforced on this info to put my mind at ease


_G__L__I__N__

[CPUShack]

They are fake for the simple reason that the CPU doesnt support any higer speeds then 333, it only has 3 pins used for determining the multiplier.
that gives 8 combinations so multipliers 4-10 (with one being used twice)

Technically it would be possible for the CPU to remap a lower multiplier but the MediaGX is not setup for that.

Not to mention you can often see the old printing under the remarked printing on the 366's (I have a 433 too lol)

[jd]

Did'nt know that , thanx.

_G__L__I__N_