Elmer, This article says Intel is now running 50nm channel lengths.
I see pictures like this everyday. I would question the 50nm from the article as accurate. It's all about where you FIB for TEM. One thing that bothers me in the pic is the Cobalt salicide formation between the top of the nitride spacer and the poly line . There is a dark region that is does not look "typical". Other items like the portion labeled nitride look familiar...a certain tool is used that creates the layered look you see in the nitride. Data is odd...you can usually gather enough data to support any conclusion you want and sell it (based on the audience)...being true / right is another issue
The article is a has a bunch of technical garble jumbled together . I guess this is the reason it's an article and not a technical paper