Power and Area-Efficient Architectural Design Methodology for Nanomagnetic Computation

Abstract

Magnetic quantum-dot cellular automata (MQCA)-based nanomagnetic logic computation started emerging to augment the CMOS-based traditional computing devices as Moore’s law approaching towards its end. Computation performed using nanomagnets exhibits non-volatility and adheres to the thermodynamic law (second). The emerging advents in the field of artificial intelligence computing on edge with the constrained resources necessitate rebooting the computing paradigm beyond CMOS and more than Moore to cater for area and power efficiency. In this regard, digital logic arithmetic circuits should be revisited using this energy-efficient computing paradigm using nanomagnets. This chapter summarizes the undergoing research in the design of such arithmetic architecture development and its corresponding nanomagnetic implementation. Researchers have demonstrated the MQCA-based arithmetic architecture implementation using inplane nanomagnetic logic (iNML) utilizing the dipole coupling. Design methodologies presented in the literatures have exploited the shape (S), positional (P), shape & positional-based hybrid nanomagnetic anisotropies pertaining to the optimization in terms of required number of resources in terms of nanomagnets (NMs), clock cycles (CCs) and majority gates (MGs) which are the critical constraints leading to high speed, area and energy-efficient design. Subsequently, researchers have exploited physical analogy of the basic building block, i.e. the three inputs nanomagnetic majority logic gate for enhanced optimization in the nanomagnetic design. However, for higher integration densities and efficient area consumption, the scalability of the dipole coupling-based nanomagnetic devices is an important aspect which is eventually limited by its susceptibility to thermal fluctuations. In this regard, interlayer exchange-coupled (IEC) scheme has been demonstrated and has been shown to offer stronger interaction between thin nanomagnets, resulting in greater scalability and better data retention at the deep sub-micron level, hence allowing magnetic interaction to be manipulated both in the vertical and lateral directions at the same time. In this regard, interlayer exchange-coupling scheme has been discussed as a possible solution to better scalability and data retention. Interlayer exchange-coupled system comprises of a non-magnetic metal layer (known as spacer layer) sandwiched between two ferromagnetic layers. The two ferromagnetic layers may be coupled ferromagnetically (FM) or antiferromagnetically (AFM), decided by the thickness and material (e.g. chromium, copper, ruthenium) of the spacer layer. On the other hand, perpendicular Nanomagnetic Logic (pNML) has involved a lot of interest for 3D architecture exploration. This chapter gives an overview on the emerging nanoscale architecture circuits, design and its implementation using nanomagnets. The implementation of nanomagnetic logic for data transmission in 3D IC has also been discussed, resulting in higher packing densities in 3D IC’s.