The International Journal Of Engineering And Science (IJES) || Volume || 5 || Issue || 9 || Pages || PP 61-68|| 2016 || ISSN (e): 2319 – 1813 ISSN (p): 2319 – 1805 www.theijes.com The IJES Page 61 Emerging Memory Technologies O.D. Alao 1 , J.V. Joshua 2 , D.O. Kehinde 3 , E.O. Ehinlafa 4 , M.O. Agbaje 5 , J.E.T Akinsola 6 1, 2, 5 ,6 Department of Computer Science, Babcock University, Ilishan-Remo, Nigeria 3 Department of Basic Science, Babcock University, Ilishan-Remo, Nigeria 4 Department of Physics, University of Ilorin, Ilorin, Nigeria. --------------------------------------------------------ABSTRACT----------------------------------------------------------- The processor caches, main memory and storage system is an integral part of any computer system. As information begins to accumulate, higher density and long term storage solutions are necessary. Due to this, computer architects face some level of challenges in developing reliable, energy-efficient and high performance memories. Also, existing storage devises are degrading in performance, cost, and sizes. Power consumption from the factory has increased, as newer codes are written, and server hardware capabilities are not adequate to handle big data of the future. New emerging memories (NEMs) are presently with its properties likely to open doors to innovative memory designs to solve the problems. This paper looks at the features of the emerging memory technologies, and compares incumbent memories types with the expected future memories. Keywords: Memory Storage, New Emerging Memory Technologies, Spin-Transfer Torque Magnetic Random- Access Memory (STT-MRAM), Resistive Random Access Memory (ReRAM) and Phase Change Memories PCM. ------------------------------------------------------------------------------------------------------------------------------------- Date of Submission: 17 May 2016 Date of Accepted: 20 September 2016 -------------------------------------------------------------------------------------------------------------------------------------- I. INTRODUCTION The pecking order of memory and storage device is a critical component of various computer systems. Processor caches act as a subset of data and instructions stored in the memory. Data stored in the main memory are stored in large, slow storage devices, such as disks and flash. Data from modern applications such books, maps, photos, audios, videos, references, facts, and conversations rely on both real and offline processing and their dataset can be in gigabytes, terabytes, zettabytes or even larger in size. Regrettably, the scaling of conventional memory technologies is at risk. Memory technologies, such as SRAM (Static Random Access Memory) and DRAM (Dynamic Random Access Memory), are experiencing scalability challenges as a result to the limitations of their device cell size and power dissipation. NEMs offer several benefits such as low power (especially low leakage), high density, and the ability to retain the stored data over long time periods (non-volatility) that have made them attractive for use as secondary storage. Flash memory is already widely used in consumer electronics and in solid-state disks due to its low cost and extremely high density [2] The dynamic and increasing power of DRAM over the power leakage of SRAM is a threat to circuit and architecture designers of future memory hierarchy designs. Energy consumption has become key design limiters as the memory hierarchy continues to contribute a significant fraction of overall system energy and power. The lack of memory technology scaling can make it difficult for the memory hierarchy to achieve high capacity and efficiency at low cost. As a result, it remains a very attractive technology for data archiving, with a sustainable roadmap for the next ten to twenty years, well beyond the anticipated scaling limits of current conventional technology [1]. There is a fundamental trend towards designing entire systems such that they are optimized for particular work- loads, departing from the traditional general-purpose architecture. The typical system, with standard CPUs consisting of a small number of identical cores with a common set of accelerators and relying on a memory and storage hierarchy has reached its limits in terms of delivering competitive performance improvements for an increasingly diverse set of workloads: future systems will be built out of increasingly heterogeneous components. This article examines today‟s memory storage requirements, reviews recent research efforts on computer architecture design with New Emerging Memories design.