Accelerator cores in low-power heterogeneous processors have on-chip local memories to enable parallel data access. The memory capacities of the local memories are very small. Therefore, the data should be transferred from the global memory to the local memories many times. These data transfers greatly increase the total processing time. Memory allocation technique to increase the data sharing is a good solution to this problem. However, when using reconfigurable cores, the data must be shared among multiple contexts. However, conventional context partitioning methods only consider how to reuse limited hardware resources in different time slots. They do not consider the data sharing. This paper proposes a context partitioning method to share both the hardware resources and the local memory data. According to the experimental results, the proposed method reduces the processing time by more than 87% compared to conventional context partitioning techniques.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Hasitha Muthumala WAIDYASOORIYA, Yosuke OHBAYASHI, Masanori HARIYAMA, Michitaka KAMEYAMA, "Memory-Access-Driven Context Partitioning for Window-Based Image Processing on Heterogeneous Multicore Processors" in IEICE TRANSACTIONS on Information,
vol. E95-D, no. 2, pp. 354-363, February 2012, doi: 10.1587/transinf.E95.D.354.
Abstract: Accelerator cores in low-power heterogeneous processors have on-chip local memories to enable parallel data access. The memory capacities of the local memories are very small. Therefore, the data should be transferred from the global memory to the local memories many times. These data transfers greatly increase the total processing time. Memory allocation technique to increase the data sharing is a good solution to this problem. However, when using reconfigurable cores, the data must be shared among multiple contexts. However, conventional context partitioning methods only consider how to reuse limited hardware resources in different time slots. They do not consider the data sharing. This paper proposes a context partitioning method to share both the hardware resources and the local memory data. According to the experimental results, the proposed method reduces the processing time by more than 87% compared to conventional context partitioning techniques.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.E95.D.354/_p
Copy
@ARTICLE{e95-d_2_354,
author={Hasitha Muthumala WAIDYASOORIYA, Yosuke OHBAYASHI, Masanori HARIYAMA, Michitaka KAMEYAMA, },
journal={IEICE TRANSACTIONS on Information},
title={Memory-Access-Driven Context Partitioning for Window-Based Image Processing on Heterogeneous Multicore Processors},
year={2012},
volume={E95-D},
number={2},
pages={354-363},
abstract={Accelerator cores in low-power heterogeneous processors have on-chip local memories to enable parallel data access. The memory capacities of the local memories are very small. Therefore, the data should be transferred from the global memory to the local memories many times. These data transfers greatly increase the total processing time. Memory allocation technique to increase the data sharing is a good solution to this problem. However, when using reconfigurable cores, the data must be shared among multiple contexts. However, conventional context partitioning methods only consider how to reuse limited hardware resources in different time slots. They do not consider the data sharing. This paper proposes a context partitioning method to share both the hardware resources and the local memory data. According to the experimental results, the proposed method reduces the processing time by more than 87% compared to conventional context partitioning techniques.},
keywords={},
doi={10.1587/transinf.E95.D.354},
ISSN={1745-1361},
month={February},}
Copy
TY - JOUR
TI - Memory-Access-Driven Context Partitioning for Window-Based Image Processing on Heterogeneous Multicore Processors
T2 - IEICE TRANSACTIONS on Information
SP - 354
EP - 363
AU - Hasitha Muthumala WAIDYASOORIYA
AU - Yosuke OHBAYASHI
AU - Masanori HARIYAMA
AU - Michitaka KAMEYAMA
PY - 2012
DO - 10.1587/transinf.E95.D.354
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E95-D
IS - 2
JA - IEICE TRANSACTIONS on Information
Y1 - February 2012
AB - Accelerator cores in low-power heterogeneous processors have on-chip local memories to enable parallel data access. The memory capacities of the local memories are very small. Therefore, the data should be transferred from the global memory to the local memories many times. These data transfers greatly increase the total processing time. Memory allocation technique to increase the data sharing is a good solution to this problem. However, when using reconfigurable cores, the data must be shared among multiple contexts. However, conventional context partitioning methods only consider how to reuse limited hardware resources in different time slots. They do not consider the data sharing. This paper proposes a context partitioning method to share both the hardware resources and the local memory data. According to the experimental results, the proposed method reduces the processing time by more than 87% compared to conventional context partitioning techniques.
ER -