ACM Home Page
Please provide us with feedback. Feedback
Scratchpad memory management in a multitasking environment
Full text PdfPdf (408 KB)
Source
International Conference On Embedded Software archive
Proceedings of the 8th ACM international conference on Embedded software table of contents
Atlanta, GA, USA
SESSION: Virtual machines, compilers, memory management table of contents
Pages 265-274  
Year of Publication: 2008
ISBN:978-1-60558-468-3
Authors
Bernhard Egger  Samsung Advanced Institute of Technology, Giheung-gu, Yongin-si, South Korea
Jaejin Lee  Seoul National University, Seoul, South Korea
Heonshik Shin  Seoul National University, Seoul, South Korea
Sponsors
ACM: Association for Computing Machinery
SIGBED: ACM Special Interest Group on Embedded Systems
SIGMICRO: ACM Special Interest Group on Microarchitectural Research and Processing
SIGDA: ACM Special Interest Group on Design Automation
Publisher
ACM  New York, NY, USA
Bibliometrics
Downloads (6 Weeks): 12,   Downloads (12 Months): 144,   Citation Count: 0
Additional Information:

abstract   references   index terms   collaborative colleagues  

Tools and Actions: Request Permissions Request Permissions    Review this Article  
DOI Bookmark: Use this link to bookmark this Article: http://doi.acm.org/10.1145/1450058.1450094
What is a DOI?

ABSTRACT

This paper presents a dynamic scratchpad memory (SPM) code allocation technique for embedded systems running an operating system with preemptive multitasking. Existing SPM allocation schemes do not support multiple tasks or only a fixed number of processes that are known at compile time. These schemes rely on algorithms that select code depending on the size of the SPM. In contemporary portable devices, however, processes are created and terminated on demand and the SPM is shared among them.

We introduce a dynamic scratchpad memory code allocation technique for code that supports dynamically created processes. At runtime, an SPM manager (SPMM) loads code pages of the running applications into the SPM on demand. It supports different sharing strategies that determine how the SPM is distributed among the running processes. We analyze several sharing strategies with regard to several preferable properties of multiprocess SPM allocation schemes.

We evaluate the proposed multiprocess SPM allocation techniques and compare them to a fully-cached reference system by running several multiprocess benchmarks. The benchmarks comprise of multiple embedded applications such as H.264, MP3, MPEG-4, and PGP. On average, we achieve a 47% improvement in throughput and a 32% reduction in energy consumption. A comparison with the unachievable lower bound shows that the best SPM sharing strategy exploits 87% of the runtime improvements and 89% of the energy savings possible.


REFERENCES

Note: OCR errors may be found in this Reference List extracted from the full text article. ACM has opted to expose the complete List rather than only correct and linked references.

1
 
2
ARM Architecture Version 6 (ARMv6). http://www.arm.com, 2002.
3
4
 
5
6
7
 
8
 
9
H.264 Video Codec. http://www.itu.int/rec/T-REC-H.264, 2003.
 
10
Intel XScale Architecture. http://www.intel.com, 2002.
 
11
12
 
13
14
 
15
16
 
17
Micron Technology, Inc. Mobile SDRAM Power Calc 10. 2004.
 
18
MP3 Reference Decoder. http://www.mp3-tech.org/programmer/sources/dist10.tgz, 1996.
19
 
20
 
21
Pretty Good Privacy (PGPi). http://www.pgpi.org/, 2002.
 
22
Samsung Semiconductor. K4X51163PC Mobile DDR SRAM. 2005.
23
24
25
 
26
M. Verma, K. Petzold, L. Wehmeyer, H. Falk, and P. Marwedel. Scratchpad sharing strategies for multiprocess embedded systems: A first approach. In 3rd Workshop on Embedded Systems for Real-Time Multimedia, September 2005.
 
27
 
28
Xvid MPEG-4 Video Codec. http://www.xvid.org, 2005.

Collaborative Colleagues:
Bernhard Egger: colleagues
Jaejin Lee: colleagues
Heonshik Shin: colleagues