Processing in Memory: Exploring the Design Space

Authors: 
M. Scrbak, M. Islam, K. Kavi, N. Jayasena and M. Ignatowski
Keywords: 
Processing-in-Memory, 3D-DRAM, Big Data, MapReduce
Abstract: 

With the emergence of 3D-DRAM, Processing-in-Memory has once more become of great interest to the research community and industry. In this paper, we present our observations on a subset of the PIM design space. We show how the architectural choices for PIM core frequency and cache sizes will affect the overall power consumption and energy eciency. Our ndings include detailed power consumption modeling for an ARM-like core as a PIM core.We show the maximum numberof PIM cores we can place in the logic layer with respect to a power budget. In addition, we explore the optimal design choices for the number of cores as a function of frequency, utilization, and energy eciency.

Publish Date: 
Tuesday, March 24, 2015
Venue: 
The 28th International Conference on Architecture of Computing Systems (ARCS-2015)
Paper URL: 
http://www.csrl.unt.edu/kavi/Research/ARCS-2015.pdf