Last month, DailyTech reported on a novel breakthrough compound, a combination of silicon and strontium titanate which when manufactured in just such a way, causes both non-ferroelectric materials to become ferroelectric. Ferroelectrics are nothing new; you can find them in various memory and "smart" cards.

So what's the hoopla all about? As the Cornell University compound is based on silicon, the de facto semiconductor in just about every piece of modern electronics, it could lead to the creation of ferroelectric transistors.

Ferroelectric compounds are handy because they can be used to create non-volatile memory. Non-volatile memory doesn't need power to save its contents. While this is all well and good, non-volatile memory isn't exactly rare either. Now if you apply that to a transistor, or the millions of transistors that make up the processors in our electronics, you might have something interesting.

These new transistors, should they come into fruition, would enable a deluge of devices known as "instant-on computers." Without the need to let an operating system button itself up for powering down, a user could simply push the power button and the computer would turn off there and then. Assuming the computer's RAM, transistors, and data storage devices were all composed of ferroelectric materials, another push of the button would bring the computer back on virtually instantly, and start it out right where it stopped - even in the middle of processing something silly like digits in pi.

Unfortunately, instant-on computers wouldn't remove the need for a boot process - the operating system and applications must still be initialized at some point - but assuming no unforeseen crashes or bugs, you would never have to reboot the machine during normal use. Simply power it off while not in use, and power it back on when you feel like shooting zombies.

In a paper lead-authored by Maitri P. Warusawithana, a process known as molecular-beam epitaxy was shown to be able to create the theoretical compound of silicon and strontium titanate. Another group at the University of Pittsburgh was able to verify by measurement that the substance was exactly what they wanted it to be, which was a ferroelectric material on a semiconductor substrate. The paper was published in last April's issue 17 of the journal Science.

Now, a paper led by Joseph Woicik of the National Institute of Standards and Technology, also a co-author of Warusawathana's paper, is showing that the compound can be used for data retention. After creating a thin film from the silicon and strontium titanate only a few molecules thick, a Cornell group led by Darrell Schlom successfully wrote, read and erased "data" to the material by a process called piezoresponse force microscopy. Though the data in this case was simple polarized domains, it shows proof of concept for all the work involved.

While the fabled instant-on silicon ferroelectric transistor still does not exist, the work being done by this long list of academies and institutions has continuously nudged it closer to reality. Rather than postulating the feasibility of existence for instant-on electronics, it may be time to think about where they will be the most useful and welcome to consumers and government.

Contributors to Woicik's paper can be found not only at NIST, but the University of Pittsburgh, Pennsylvania State University, Northwestern University, Motorola, Ames Laboratory at the Department of Energy, Intel Corporation and Tricom Tech with data taken from the Advanced Photon Source at Argonne National Laboratory. The work has thus far been funded by the National Science Foundation and the Office of Naval Research.