Energy-Efficient Near-Threshold Parallel Computing: The PULPv2 Cluster

This article presents an ultra-low-power parallel computing platform and its system-on-chip (SoC) embodiment, targeting a wide range of emerging near-sensor processing tasks for Internet of Things (IoT) applications. The proposed SoC achieves 193 million operations per second (MOPS) per mW at 162 MO...

Full description

Saved in:
Bibliographic Details
Published in:IEEE MICRO 2017-09, Vol.37 (5), p.20-31
Main Authors: Rossi, Davide, Pullini, Antonio, Loi, Igor, Gautschi, Michael, Gurkaynak, Frank Kagan, Teman, Adam, Constantin, Jeremy, Burg, Andreas, Miro-Panades, Ivan, Beigne, Edith, Clermidy, Fabien, Flatresse, Philippe, Benini, Luca
Format: Article
Language:English
Subjects:
body biasing
Computation
Energy efficiency
Internet of Things
Low power electronics
Memory management
Mopping
parallel processing
power management
Power system management
Random access memory
Reduced instruction set computing
System on chip
UTBB FD-SOI
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This article presents an ultra-low-power parallel computing platform and its system-on-chip (SoC) embodiment, targeting a wide range of emerging near-sensor processing tasks for Internet of Things (IoT) applications. The proposed SoC achieves 193 million operations per second (MOPS) per mW at 162 MOPS (32 bits), improving the first-generation Parallel Ultra-Low-Power (PULP) architecture by 6.4 and 3.2 times in performance and energy efficiency, respectively.
ISSN:0272-1732
1937-4143
DOI:10.1109/MM.2017.3711645