D. K. Karunaratne

Study of single layer and multilayer nano-magnetic logic architectures

Nano-magnetic logic (NML) has been a promising technology for logic computation. Contribution of this paper is two-fold. First, we have fabricated and captured MFM images of a NML architecture that has computed the majority of seven variables. This logic block can potentially implement eight different logic functions that could be configured in real-time. Next, we have performed a set of experiments with a multilayer stack of Cot = 0.4 nm/Cut = 4 nm/Ni20Fe80t = 5 nm with a perpendicular magnetic anisotropy bottom layer to realize neighbor interaction between adjacent free layers of devices. Based on the MFM images, we conclude that dipolar coupling between the free layers of the neighboring spin-valve based NML (SVBN) devices can be exploited to construct local elements such as majority gates, inverters and interconnects. Since magnetic multilayer stacks have already been implemented in memory devices to read/write data, SVBN devices would not only solve the input/output problems in NML but also would hav...

Non-Boolean computing with nanomagnets for computer vision applications

The field of nanomagnetism has recently attracted tremendous attention as it can potentially deliver low-power, high-speed and dense non-volatile memories. It is now possible to engineer the size, shape, spacing, orientation and composition of sub-100 nm magnetic structures. This has spurred the exploration of nanomagnets for unconventional computing paradigms. Here, we harness the energy-minimization nature of nanomagnetic systems to solve the quadratic optimization problems that arise in computer vision applications, which are computationally expensive. By exploiting the magnetization states of nanomagnetic disks as state representations of a vortex and single domain, we develop a magnetic Hamiltonian and implement it in a magnetic system that can identify the salient features of a given image with more than 85% true positive rate. These results show the potential of this alternative computing method to develop a magnetic coprocessor that might solve complex problems in fewer clock cycles than traditional processors.

Study of Dipolar Neighbor Interaction on Magnetization States of Nano-Magnetic Disks

We investigate the effect of magnetic neighbor interaction on the state behavior of nano-magnetic disks for data storage and computation applications. We have observed and verified that a nano-magnetic disk, with certain dimension, can exist either in the single domain state or in the vortex state depending on the edge-to-edge spacing between the nano-magnetic disks. The experiments were conducted by varying the diameters and thicknesses with respect to edge-to-edge spacing. The dimensions were based on the phase diagram between the single domain state and the vortex state. We have observed nano-magnetic disks spaced far apart from its neighbor, settled in the vortex state and coupled nano-magnetic disks with less spacing settled in the single domain state. This phenomenon was observed for nano-magnetic disks with thickness between 8 nm to 20 nm and diameters between 80 nm to 140 nm. We validated the simulation by fabricating pairs of nano-magnetic disks with a thickness of 10 nm and diameter of 110 nm along edge-to-edge spacing from 20 nm to 260 nm in steps of 20 nm.

Study of magnetization state transition in closely spaced nanomagnet two-dimensional array for computation

The present study investigated the dipole–dipole interaction for finite 2D arrays of ferromagnetic circular nanomagnets. Starting with two basic arrangements of coupled nanomagnets namely, longitudinal and transverse, different diameters, and thicknesses are studied. The phase plot results exhibit for longitudinal arrangements that the single domain state is pervasive over a large range of thickness values as compared to the transverse arrangement or isolated nanomagnet cases. The study is further extended to finite arrays (3 × 3 and 5 × 5) of circular nanomagnets. The magnetic force microscopy results show that arrays of nanomagnets favors antiferromagnetic ordering at remanence. We have correlated our experimental results with micromagnetic simulations. Based on our study, we can conclude that nanomagnets with 100 nm in diameter, 15 nm in thickness, and 20 nm in spacing have single domain states in an array configuration with one-step switching, which results in fast operation, a property ideal for comp...

Magnetic State Estimator to Characterize the Magnetic States of Nano-Magnetic Disks

We present an image processing system-magnetic state estimator (MSE) to estimate the magnetization states of lithographically patterned single domain nano-magnets. The tool requires digital images of a magnetic force micrograph, an atomic force micrograph and a computer-aided design layout of the magnetic nanostructures to estimate their magnetization states. The tool also yields a confidence value of the estimated magnetization state. We have tested the MSE on randomly patterned Permalloy nano-magnetic disks on a Silicon wafer. The tool was able to correctly identify 403 single domain states out of 499 and 256 vortex states out of 297, producing an accuracy rate of 83%.

Driving magnetic cells for information storage and propagation

Single domain magnets serve as an excellent mechanism to store and preserve Boolean information. This work is an effort to establish that information can not only be stored and altered as in memory but also can be propagated in a causal fashion from the driver to the driven cell in presence of an external clocking field.

Novel knowledge module on fusion of logic and memory to undergraduate students

This work is an effort to introduce short knowledge modules targeting new application or emerging devices in conventional “core” courses. In this knowledge module, we report introduction of two emerging devices, Memristor and Magnetic Cellular Automata. Both the devices also have potential to work as memory and logic simultaneously. Based on the worksheet performance, we observed that students were excited, motivated and assimilated the key components of the technical module. We also conducted an external independent survey, which suggest significant positive perception, interest towards the novel logic paradigm in the post-module survey compared to the pre-module survey.

Tool for analysis and quantification of fabrication layouts in nanomagnet-based computing

The coupling among nano-magnets, placed at specific locations, but not necessarily in a regular grid, can be used to perform logic operations. We have developed an image-processing tool that will help one to automatically analyze and characterize atomic force microscopy (AFM) images and magnetic force microscopy (MFM) images of fabrication attempts to identify errors and magnetization. This is currently a highly manual and tedious process. We quantify fabrication errors and generate an enhanced MFM image. The fabrication errors are quantified into 3 categories: missed, extra and merged. The enhanced MFM image shows only magnetization of the sample without any surface characteristics.