RESEARCH TO COMPREHEND NEXT-GENERATION COMPUTER MEMORY ARCHITECTURES BY USING MODELLING TECHNIQUES AND UTILISING EMERGING NON-VOLATILE MEMORIES

Abstract

Modern computer system design places a premium on energy efficiency. Since conventional CMOS scaling theory states that threshold and supply voltages are reduced in relation to device sizes, the widespread assumption is that leakage would grow exponentially with shrinking CMOS technology. Therefore, modern techniques count leaky power as a competitor to dynamic power. The power budget leakage problem can't be solved unless there's a surge of innovative, game-changing technology. A number of noteworthy new developments have occurred in the area of non-volatile memory technology. Popular examples of contemporary non-volatile memories with desirable characteristics such as low access energy, high cell compactness, and outstanding access performance include "ReRAM," "PCRAM," and "Spin-Torque-Transfer Random Access Memory" (MRAM, STTRAM). So, it's fantastic that these new non-volatile memory technologies will be used to construct future computers that are both powerful and energy efficient. To prove their value, further academic study is needed, since these new non-volatile memory technologies are still in the research and development phase. Because of this, three methods for facilitating these new forms of non-volatile memory are explored in this research. We begin with models of several forms of nonvolatile memory, including their space requirements, power consumption, and performance at the circuit level. Second, they assess the effects of write operations on non-volatile memory and provide several techniques at the architectural level to reduce such effects. Finally, they look at actual uses of this cutting-edge innovation in case studies.