홍혜정, Hyun Jin Kim, 강성호
Abstract
Reconfigurable coprocessors take charge of time-consuming and data-intensive computations. Speedup of reconfigurable coprocessors can lead performance improvement of entire systems. In this paper, we propose a new scheduling and mapping method considering the characteristics of applications. The achieved speedup ranges from 11% to 35%.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Reconfigurable coprocessors take charge of time-consuming and data-intensive computations. Speedup of reconfigurable coprocessors can lead performance improvement of entire systems. In this paper, we propose a new scheduling and mapping method considering the characteristics of applications. The achieved speedup ranges from 11% to 35%.
Key concepts: Speedup, Coprocessor, Computer science, Parallel computing, Scheduling (production processes), Computation, Algorithm, Mathematics