Intel's 1.9Bn Transistor, 32nm Die Works

Loading...

Been waiting a while? Get Flash to see this player.

Sanjay Natarajan is driving Intel's efforts toward 32nm CPUs for 2009, setting up a key process battle between Intel and AMD. While team-green are betting the farm on Low-K Dielectric Technology, Intel say High-K is the way to go...

Related Reading


HEXUS Forums :: 18 Comments

Login with Forum Account

Don't have an account? Register today!
Absolutely fascinating !

AMD say Low-K is the future

Intel say no way, choose High-K

But who is right ?

If I understand this correctly, then AMD are saying that by putting in traffic calming measures, with speed bumps and a reduction in ‘alternate path short cuts’ for errant electrons - the world will be a better place…

…while Intel prefer to make the highway connecting 2 items much faster - so that any deviation is automatically discouraged because electrons follow the path of least resistance

(if I have the fuzzy end of this lolly-pop, then please feel free to jump in and correct me!)

Both are on schedule to deliver mass-market volume on next-gen parts by 2009/10 and then - I guess - the stock market will vote with its wallet

.
Thats pretty cool.

Surely both AMD and Intel could be right with their differing approaches to metal gates? They are just approaching the problem from different angles.
Can't they use high-K for some part of the chip and low-K for another part of the chip?
I'm not sure you can mix it due to it being a manufacturing process, I'm no expert on silicon manipulation but I'm confident that it would take a ‘double pass’ on the line and you would need to mask the other area. I will find out though
Andrzej
AMD say Low-K is the future

Intel say no way, choose High-K

But who is right ?


Kellogs and their Special K ;)

As long as both the processes work as intended then they can use any they want :)