A group of boffins from the Applied Electromagnetics Group at the University of California San Diego have succeeded in building a microelectronic device without the incorporation of semiconductors for the first time at least in a published paper. It’s certainly possible that the feat has been accomplished in a secret military facility or by a want to be Bond villain, but this research was published this week.
Fighting Moore’s Law by abandoning semiconductors
The team published its findings in Nature Communications this week explaining how they were able to engineer an optically-controlled device that bypassed by the use of semiconductors by thinking of and employing a metasurface in the nanoscale that avoids the limits that superconductors bring to electron flow or, perhaps more familiar, conductivity. Electrons have a pesky habit of running into some atoms on their way to a given point in a semiconductor, but the team worked out a vacuum tube 2.0 that avoids the limitations placed on power handling and speed that semiconductors bring.
Work with my layman understanding of how this was accomplished with the knowledge that I’m working from my own brain’s limitations when quantum or nano understandings are brought to the table in spite of near prodigy level mathematical understandings I once possessed but aren’t challenged terrifically before being blown away by these innovations.
The following is our rough coverage of the 2021 Sohn Investment Conference, which is being held virtually and features Brad Gerstner, Bill Gurley, Octahedron's Ram Parameswaran, Glenernie's Andrew Nunneley, and Lux's Josh Wolfe. Q1 2021 hedge fund letters, conferences and more Keep checking back as we will be updating this post as the conference goes Read More
By using a less than 10 volt DC charge and infrared laser on a metasurface the team made to rest on a silicon wafer with a layer of silicon dioxide, they found they could make their chip work. On a nanoscale, the researchers couldn’t simply apply high powered lasers that would have replaced the need for the metasurface used to boost conductivity owing to the heat that lasers would produce.
“This certainly won’t replace all semiconductor devices, but it may be the best approach for certain specialty applications, such as very high frequencies or high power devices,” said electrical engineering professor Dan Sievenpiper at UC San Diego’s Applied Electromagnetics Group who led the team and was the lead author of the paper published this week.
“Next we need to understand how far these devices can be scaled and the limits of their performance.”
Moore’s Law dictates that computing speeds double every two years and are meant to see a consequent drop in cost and the team recognizes that they are on to something impractical beyond their existing accomplishment.
We’re years, if not multiple decades away from the replacement of the semiconductors that power more and more of our lives each day. We still see silicon pushing the boundaries of its capacity. Thinner and faster will still be the way forward as power needs decrease, but ionic traps and the programming of the atom, only discovered a touch over a century ago, are the future and these kids are smart.
The work done by this team show that the use of low-power lasers and DC power clearly illustrate that when applied to the correct surfaces can replace the need for a high-powered laser, but scalability and further advances will have to wait another five years(?). When Elon Musk presented his white paper on the Hyperloop he was scoffed at, but others have run with this, and it shows the power of outside the box thinking with talented individuals doing the thinking.