Half-a-Loaf Method Can Improve Magnetic Memories

Image by Getty Images via @daylifeChinese scientists have shown that magnetic memory, logic and sensor cells can be made faster and more energy efficient by using an electric, not magnetic, field to flip the magnetization of the sensing layer only about halfway, rather than completely to the opposite direction. They describe the new cell design in the Journal of Applied Physics, which is published by the American Institute of Physics (AIP).

Magnetic random access memory (or MRAM) cells have long been investigated as possible replacements for parts of hard disk drives, flash memory and even computing circuits. Previous designs, however, have proven to be too power-hungry or expensive to be competitive.

"Our new cell design offers a great possibility for data storage elements and logic gates that are fast and non-volatile with ultra-low power consumption," said Dr. Ce-Wen Nan of Tsinghua University in Beijing, China. The new cell is also simpler to make than existing components. Only two layers are needed, compared with three or more for traditional magnetic memories.

The design by Nan's group is a simple thin-layer sandwich of two different materials, each of which has very different magnetic and electrical properties. Applying a voltage to the ferroelectric layer switches its polarization in a way that starts to change the magnetic orientation of the adjacent ferromagnetic layer. This partial change alters the electrical resistance of the entire stack enough to indicate whether the cell is storing a "0" or a "1" data bit. Future research is aimed at understanding and optimizing the materials to increase the resistance change, which will enhance its commercial prospects.

Microsoft Excel-Based Algorithm Predicts Cancer Prognosis

Image via Wikipedia

Robin Hallett, a graduate student working under the supervision of Dr. John Hassell and other members of his research team from McMaster University, Ontario, Canada, developed the algorithm and used it to identify a 20 gene signature, which performed well on a 151 patient validation dataset.

Hallett said, "Until now, constructing such a signature requires the use of various clustering and classification algorithms, which in turn require specialized software and bioinformatics training. Importantly, we completed all steps of our algorithm using Microsoft Excel 2007. This software is widely, if not universally, accessible to the biological research community, suggesting that implementation of this technique will not be hampered by lack of software or training."

The researchers used data from a group of 144 patients to train the algorithm to identify genes whose expression levels correlated with patient survival. The 10 most highly ranked genes predictive of poor prognosis and those 10 genes most highly predictive of good prognosis established a 20-gene expression based predictor, which was found to perform as well as two other models in the validation group.

According to Hassell, "Our algorithm produces prediction models with comparable accuracy to other feature selection techniques while having generally better accessibility and useability for biological research scientists. We've begun using our algorithm to generate gene expression based prediction models of breast cancer cell sensitivity to commonly used anti-cancer therapies."