Non-Traditional Chips Gaining Steam

Question for the collective:<p>As data rates in cables have gone up, designers have resorted to putting codecs into the housing at the end(s) of the cables to keep up. In some cases this has made for connectors that are only slightly smaller than those from the 90’s.<p>Can flexible circuits switch at a high enough rate to perform these signal processing tasks? I’m imagining building the circuits into the ends of the cable itself (under the strain protector?) instead of the housing. Or do we just wait for smaller chips?

“The whole market is getting closer to reality,” Freeman says. “I don’t think substrates are a gating factor. We’re still looking for that killer app.”<p>How about an inkjet printer that can print OLEDs and circuits? People can design light-up signs and even have small circuits to control changing patterns. I want to print a processor with RAM&#x2F;ROM on one side and some kind of display on the other. This would be both widely usable and a boon to electronics education.

Can someone help me understand - is the non-traditionalness of these chips enabled by continued shrinking of transistors, or something else?<p>Are we on a new learning curve?