American boffins say they have made a significant step towards new chip technologies which could lead to faster, more durable non-volatile memory in place of RAM - and so to "instant-on" computing without boot-ups or hard disks.

The new research comes in the area of ferroelectrics, materials used today in so-called FeRAM or FRAM* - Ferroelectric Random Access Memory. FeRAM is a type of non-volatile memory used in smart cards and some portable devices. Compared to Flash, it offers many more read/write cycles before failure, much faster writing and lower power consumption.

FeRAM chips are structured much like ordinary DRAM, but use a ferroelectric material in place of the dielectric to achieve their unique features. Present-day FeRAM uses lead zirconate titanate (PZT) ferroelectric, and at present the technology suffers from low storage densities and high cost: hence its use mainly in low-capacity low-power devices requiring durability, like smartcards.

But now boffins funded by the US National Science Foundation have made a move to change that, depositing strontium titanate on silicon to achieve a new ferroelectric.

According to the NSF, this could lead to new, superior FeRAM chips able to replace both the RAM and the hard-disk or Flash storage used in most present-day computers. The FeRAM combo-memory would of course be non-volatile, unlike today's RAM, so there'd be no need to boot up or recover from hibernation after turning off the power: the machine would instantly switch on in just the same condition it had been on power-down. If the FeRAM also replaced the machine's storage, there'd never be any tiresome waiting for slow, power-hungry hard disks or Flash memory to do their stuff either.

The research was led by Darrell Schlom of Cornell Uni, with assistance from federal boffins - including some from NASA Ames - and others at Motorola and Intel.

It's not goodbye to RAM just yet, however, according to Schlom. He and his collaborators have shown that strontium-titanate components can be assembled, but there's still plenty to do.

"Several hybrid transistors have been proposed specifically with ferroelectrics in mind," said Schlom. "By creating a ferroelectric directly on silicon, we are bringing this possibility closer to realization."

Subscribers to the journal Science can read the paper here. There's also some background on FeRAM tech from Fujitsu here, for those interested. ®

*FRAM is specifically trademarked by Ramtron International.