RAINBOW: Multi-Dimensional Hardware-Software Co-Design for DL Accelerator On-Chip Memory
Paper in proceeding, 2023

Deep Learning (DL) is developing at an extremely fast pace. The increased number of applications, optimizations and hardware devices available, results in a multi-dimensional design space where the best performance is achieved with a detailed analysis of the hardware-software co-design process. Furthermore, the high demands for memory and the off-chip latency cost result in the on-chip memory becoming critical for achieving high performance and efficiency. In this work, we propose RAINBOW, a tool to assist in the hardware-software co-design for DL accelerators' on-chip memory. The purpose is to help the design and/or deployment of a DL model to a dedicated accelerator. RAINBOW generates different analyses results and feeds them to the optimizers. The result is a heterogeneous execution plan combining different approaches and techniques depending on the dynamic requirements and constraints. In our analysis, we concluded that given the opportunity, RAINBOW'S heterogeneous plans are able to reduce the DRAM accesses to approximately half when compared to homogeneous plans.

multi-dimensional design

Deep Learning

Author

Stavroula Zouzoula

Chalmers, Computer Science and Engineering (Chalmers), Computer Engineering (Chalmers)

Muhammad Waqar Azhar

Chalmers, Computer Science and Engineering (Chalmers), Computer Engineering (Chalmers)

Pedro Petersen Moura Trancoso

Chalmers, Computer Science and Engineering (Chalmers), Computer Engineering (Chalmers)

Proceedings - 2023 IEEE International Symposium on Performance Analysis of Systems and Software, ISPASS 2023

352-354
9798350397390 (ISBN)

2023 IEEE International Symposium on Performance Analysis of Systems and Software, ISPASS 2023
Raleigh, USA,

Very Efficient Deep Learning in IOT (VEDLIoT)

European Commission (EC) (EC/H2020/957197), 2020-11-01 -- 2023-10-31.

Subject Categories

Computer and Information Science

Computer Systems

DOI

10.1109/ISPASS57527.2023.00050

More information

Latest update

8/14/2023