A Programmable Hyper-Dimensional Processor Architecture for Human-Centric IoT
- 15 August 2019
- journal article
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Emerging and Selected Topics in Circuits and Systems
- Vol. 9 (3) , 439-452
Abstract
Hyper-dimensional Computing (HDC), a bioinspired paradigm defined on random high-dimensional vectors, has emerged as a promising IoT paradigm. It is known to provide competitive accuracy on sequential prediction tasks with much smaller model size and training time compared to conventional ML, and is well-suited for human-centric IoT. In the post-Moore scaling era, where increasing variability has challenged traditional designers, its novel computing method based on randomness can be leveraged for continued performance. This work develops a complete, programmable architecture for ultra energy-efficient supervised classification using HD computing. Its simple construction follows from basic HD operations and its massively parallel, shallow datapath (¡ 10 logic layers) resembles in-memory computing. The architecture also supports scalability: multiple such processors can be connected pralallely to increase effective HD dimension. A broad evaluation is performed by comparing HDC and 3 conventional ML algorithms on conventional architectures such as CPU and eGPU for instruction count, energy cost and memory requirements. Finally, a 2048-dim ASIC design is synthesized in a 28nm HK/MG process and benchmarked on 9 supervised classification tasks with varying complexity (such as language recognition and human face detection). The simulated chip exhibits energy efficiency ¡ 1.5μJ/pred. for the entire benchmark at about 2.5ns cycle time, with most applications requiring ¡ 700 nJ/pred. As a first complete design working with high dimensional stochastic signals, the main architectural decisions for similar systems harnessing variability in emerging devices (eg. CNFET and RRAM) are established. A fabricated system could be readily deployed for human-centric IoT applications.Keywords
Funding Information
- National Science Foundation
This publication has 53 references indexed in Scilit:
- LIBSVMACM Transactions on Intelligent Systems and Technology, 2011
- Classification through incremental max–min separabilityPattern Analysis and Applications, 2010
- Challenges for Nanoscale MOSFETs and Emerging NanoelectronicsTransactions on Electrical and Electronic Materials, 2010
- A study of the behavior of several methods for balancing machine learning training dataACM SIGKDD Explorations Newsletter, 2004
- Hierarchical algorithm partitioning at system level for an improved utilization of memory structuresIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 1999
- A processor-time-minimal systolic array for cubical mesh algorithmsIEEE Transactions on Parallel and Distributed Systems, 1992
- Synthesis of control circuits in folded pipelined DSP architecturesIEEE Journal of Solid-State Circuits, 1992
- Array architectures for block matching algorithmsIEEE Transactions on Circuits and Systems, 1989
- A family of new efficient arrays for matrix multiplicationIEEE Transactions on Computers, 1989
- Some inequalities relating to the partial sum of binomial probabilitiesAnnals of the Institute of Statistical Mathematics, 1959