Intel 10nm COAC broken?

By: Brett (, November 25, 2018 7:42 pm
Room: Moderated Discussions
Intel 10nm uses Contact Over Active Gate (COAG), is this fundamentally broken?


My speculation is that the contact is proving a thermal conductor resulting in uneven heating of the transistor, which in turn leads to expansion stress on the oxide layer that could lead to progressive atomic slippage and fracturing, leading to failure.

This area is poorly modeled, leading to wild goose searches for the failures and a fix.

The fix could be not using COAG.
But that means re-layout of the entire chip, which for Intel means a year and 100 million dollars, and it could take a year before the fab guys give up on trying to get COAG working.

Then there is the issue of creating a team to re-layout the chip, when the original team has a year before moved onto the next project.
The internal politics of such a disaster are epic, and would explain how Intel fell behind the industry in shrinks, as management instead kept saying just fix the existing COAG.

The tiny sample of 10nm Intel CPU chip sneaked onto the market without a working graphic part and running at low clocks may be a test of a COAG-less chip, the graphics part is not ported to the newer 10nm process yet?
 Next Post in Thread >
TopicPosted ByDate
Intel 10nm COAC broken?Brett2018/11/25 07:42 PM
  Intel 10nm COAC broken?blue2018/11/25 09:12 PM
    COAG is brokenDavid Kanter2018/11/25 11:52 PM
    Intel 10nm COAC broken?Brett2018/11/26 01:00 AM
      Intel 10nm COAC broken?Simon Farnsworth2018/11/26 02:27 AM
      Intel 10nm COAC broken?Maynard Handley2018/11/26 12:24 PM
  Intel 10nm COAC broken?Groo2018/11/26 05:35 AM
    Intel 10nm COAC broken?Youwillnothavemymoney2018/11/26 05:53 AM
      Intel 10nm COAC broken?Doug S2018/11/26 10:20 AM
      10 nm is currently really unprofitableJames2018/11/28 03:30 AM
Reply to this Topic
Body: No Text
How do you spell purple?