IBM may have extended the lifespan of Moore’s Law

Awful article. Clueless writer. Better article from the Register.[1]

Actual news: IBM fabs test chip with 7nm line width. Intel is at 10nm right now. IBM had to use "extreme ultraviolet" and silicon germanium rather than plain silicon. Both are known technologies. IBM is spending $3 billion. Everyone has been trying to avoid going to "extreme ultraviolet", which is really soft X-rays. They're produced by a synchrotron or some exotic laser tricks, not a lamp, and the cost of the "light source" is high.

The other side of Moore's Law is that each generation of fab costs a lot more than the last one. Also, getting rid of the heat now dominates; only mostly-idle memory devices like flash reap the full benefit of the highest densities.

For reference, a single atom is about 0.1nm. Experimentally, wires 1 atom wide and 4 atoms high have been made. A single-atom transistor is theoretically possible. There's still room at the bottom.

[1] http://www.theregister.co.uk/2015/07/09/ibm_cracks_7nm_barri...

> According to IBM, this could lead to a 50% performance and power boost over chips that are on the market today, effectively keeping Moore’s Law more or less intact for the time being.

I propose when we finally break the 'law' of Moore's Law, that's when we should say it's not intact. All this "we're at the limit" stuff is just hot air, IMHO.

Compute is going to continue to increase at a steady rate and networking is going to lag behind just a little bit less. That's the real mess in all this, not Moore's Law breaking!