An Architecture for Memory Centric Active Storage (MCAS)

The advent of CPU-attached persistent memory technology, such as Intel’s Optane Persistent Memory Modules (PMM), has brought with it new opportunities for storage. In 2018, IBM Research Almaden began investigating and developing a new enterprise-grade storage solution directly aimed at this emerging technology. MCAS (Memory Centric Active Storage) defines an “evolved” network-attached key-value store that offers both near-data compute and the ability to layer enterprise-grade data management services on shared persistent memory. As a converged memory-storage tier, MCAS moves towards eliminating the traditional separation of compute and storage, and thereby unifying the data space. This paper provides an in-depth review of the MCAS architecture and implementation, as well as general performance results.

[1]  Darrell D. E. Long,et al.  Optimizing Systems for Byte-Addressable NVM by Reducing Bit Flipping , 2019, FAST.

[2]  Andrew S. Tanenbaum,et al.  Distributed systems: Principles and Paradigms , 2001 .

[3]  Tajana Simunic,et al.  Exploring Processing In-Memory for Different Technologies , 2019, ACM Great Lakes Symposium on VLSI.

[4]  Maurice Herlihy,et al.  Hopscotch Hashing , 2008, DISC.

[5]  Ran Ginosar,et al.  Resistive GP-SIMD Processing-In-Memory , 2016, ACM Trans. Archit. Code Optim..

[6]  Tajana Simunic,et al.  FloatPIM: In-Memory Acceleration of Deep Neural Network Training with High Precision , 2019, 2019 ACM/IEEE 46th Annual International Symposium on Computer Architecture (ISCA).

[7]  James Reinders,et al.  Intel® threading building blocks , 2008 .

[8]  Xiao Liu,et al.  Basic Performance Measurements of the Intel Optane DC Persistent Memory Module , 2019, ArXiv.

[9]  Daniel Waddington,et al.  Evaluating Intel 3D-Xpoint NVDIMM Persistent Memory in the Context of a Key-Value Store , 2020, 2020 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS).

[10]  Shyh-Jye Jou,et al.  On EDA Solutions for Reconfigurable Memory-Centric AI Edge Applications , 2020, 2020 IEEE/ACM International Conference On Computer Aided Design (ICCAD).