Replies to post #15922 on Intel Corp (INTC)

[Elmer Phud]

mike

Let me refrase the main point I wanted to make. After IMB clearly demonstrated the goodness of Cu wires as compared to Al wires, nobody wanted to jump into it. In fact, if you recall, initially Intel sais that's nonsense. However as soon as AMD demonstrated that Cu is the way to go, most of the others got into it soon thereafter.

Intel never said Cu was nonsense. Intel didn't need Cu for their .18u process so they didn't use it. AMD needed Cu @.18u to be competitive, that's why they used it first.

So my main point was that, after Intel's recent demonstration, mostly everyone will jump at it, and invest in Si photonics. This will speed up the development. Do you agree?

I'm sure Intel has the proper patents in the works. Clearly there is no prior art. If Intel can actually productize this then it has the potential to be the equivalent to a patent on the transistor. However it's a long way from the lab to a product you can manufacture at a reasonable cost with both performance and reliability. I don't see it in less than 8-10 years, if then. I'd be delighted to be wrong.

You are right! 65 nm is already here, but (from a practical point of view) it's hard to believe it will go any lower. Deep UV lithography can't be pushed to much more (even if bottom layer antireflection coatings are used), and the electron lithography (and other fancy solutions) as I know is still not usable.

Well process is not my field but Intel seems to believe they have at least 10 years left and I tend to believe them.

The stakes are too high, and whoomever doesn't invest in it (IO's) will be left behind. The ones that accumulate most of IP's in this area have the biggest chance to conquer a good slize of the future market. Since I'm impressed with Intel's results (after only some 2 years), most probably (or at least is what I hope) they will keep the lead. But, once they "oppened their mouth," they better move fast.

I agree that this is a very exciting area and if Intel can corner the IP then there's no telling what this could mean.

[Tenchu]

Mike, So my main point was that, after Intel's recent demonstration, mostly everyone will jump at it, and invest in Si photonics. This will speed up the development. Do you agree?

Photonics is a whole different beast than copper interconnect. Your statement is akin to saying, "Well, we broke the sound barrier, so breaking the light speed barrier is just a generation away."

Tenchu