Further GPU acceleration of predictive partitioned vector quantization for ultraspectral sounder data compression

Shih-Chieh Wei; Bormin Huang

doi:10.1117/12.894390

16 September 2011 Further GPU acceleration of predictive partitioned vector quantization for ultraspectral sounder data compression

Shih-Chieh Wei, Bormin Huang

Author Affiliations +

Proceedings Volume 8157, Satellite Data Compression, Communications, and Processing VII; 815704 (2011) https://doi.org/10.1117/12.894390
Event: SPIE Optical Engineering + Applications, 2011, San Diego, California, United States

Abstract

For the ultraspectral sounder data which features thousands of channels at each observation location, lossless compression is desirable to save storage space and transmission time without losing precision in retrieval of geophysical parameters. Predictive partitioned vector quantization (PPVQ) has been proven to be an effective lossless compression scheme for ultraspectral sounder data. It consists of linear prediction, bit-depth partitioning, vector quantization, and entropy coding. In our previous work, the two most time consuming stages of linear prediction and vector quantization were identified for GPU implementation. For GIFTS data, using a spectral division strategy for sharing the compression workload among four GPUs, a speedup of ~42x was achieved. To further enhance the speedup, this work will explore a spatial division strategy for sharing workload in processing the six parts of a GIFTS datacube. As result, the total processing time of a GIFTS datacube on four GPUs can be less than 13 seconds which is equivalent to a speedup of ~72x. The use of multiple GPUs for PPVQ compression is thus promising as a low-cost and effective compression solution for ultraspectral sounder data for rebroadcast use.

Citation Download Citation

Shih-Chieh Wei and Bormin Huang "Further GPU acceleration of predictive partitioned vector quantization for ultraspectral sounder data compression", Proc. SPIE 8157, Satellite Data Compression, Communications, and Processing VII, 815704 (16 September 2011); https://doi.org/10.1117/12.894390

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available