How reliable are first version processors?

[Mark R]

Let's imagine that I'm an engineer at a CPU manufacturer.

The very first prototype of a new CPU has just come back from the research fab, along with a brand new chipset, and a reference mobo.

I hook it all up and add the requisite bits. Would it be stable enough to boot up to, say, an Windows XP desktop?

I'm guessing that with all the simulations, etc. that it probably should be able to - but how much debugging and testing has to wait for silicon?

How long, roughly, does it take to go from first silicon to a commercial product?

 

[otispunkmeyer]

Originally posted by: Mark R
Let's imagine that I'm an engineer at a CPU manufacturer.

The very first prototype of a new CPU has just come back from the research fab, along with a brand new chipset, and a reference mobo.

I hook it all up and add the requisite bits. Would it be stable enough to boot up to, say, an Windows XP desktop?

I'm guessing that with all the simulations, etc. that it probably should be able to - but how much debugging and testing has to wait for silicon?

How long, roughly, does it take to go from first silicon to a commercial product?

not sure, i heard it takes 3 months to go from first tape out to commercial product with GPU's but im really not sure

 

[kd2777]

First thing I would do is set the FSB to 245 and then boot up.

j/k

Interested in this too. Marked for later.

kd

 

[kravmaga]

after they run memtest and prime95 they run 3Dmark06 and use elite hacking skills.