Write a Blog >>
LCTES 2019
Sat 22 - Fri 28 June 2019 Phoenix, Arizona, United States
co-located with PLDI 2019
Sun 23 Jun 2019 09:45 - 10:00 at 105A - Session 1: Memory Management Chair(s): Greg Stitt

Tensor contraction is a fundamental operation in many algorithms with a plethora of applications ranging from quantum chemistry over fluid dynamics and image processing to machine learning. The performance of tensor computations critically depends on the efficient utilization of on-chip memories. In the context of low-power embedded devices, efficient management of the memory space becomes even more crucial, in order to meet energy constraints. This work aims at investigating strategies for performance- and energy-efficient tensor contractions on embedded systems, using racetrack memory (RTM)-based scratch-pad memory (SPM). Compiler optimizations such as the loop access order and data layout transformations paired with architectural optimizations such as prefetching and preshifting are employed to reduce the shifting overhead in RTMs. Experimental results demonstrate that the proposed optimizations improve the SPM performance and energy consumption by 24% and 74% respectively compared to an iso-capacity SRAM.

Sun 23 Jun
Times are displayed in time zone: (GMT-07:00) Tijuana, Baja California change

09:45 - 11:00: LCTES 2019 - Session 1: Memory Management at 105A
Chair(s): Greg StittUniversity of Florida
LCTES-2019-papers09:45 - 10:00
Asif Ali KhanTU Dresden, Germany, Norman A. RinkTU Dresden, Germany, Fazal HameedTU Dresden, Germany, Jeronimo CastrillonTU Dresden, Germany
LCTES-2019-papers10:00 - 10:15
LCTES-2019-papers10:15 - 10:30
Thomas Haywood DadzieHanyang University, Jiwon LeeHanyang University, Jihye KimKookmin University, Hyunok OhHanyang Univ
LCTES-2019-papers10:30 - 10:45
Yejia DiEast China Normal University, Chongqing University, Liang ShiChongqing University, Shuo-Han ChenAcademia Sinica, Taipei, Jason XueCity University of Hong Kong, China, Edwin ShaChongqing University
LCTES-2019-papers10:45 - 11:00
Chundong WangSingapore University of Technology and Design, Sudipta ChattopadhyaySingapore University of Technology and Design, Gunavaran BrihadiswarnUniversity of Moratuwa