Efficient video processing at scale using MSVP

H. M. Reddy; Y. Chen; J. Lan; I. Katsavounidis; B. Anandharengan; H. G. Lalgudi; S. Alaparthi; G. Hua; H.-C. Chuang; P.-H. Wu; Z. Lei; A. Mastro; C. Petersen; G. Chaudhari; P. Prakash; S. Regunathan; S. Reddy; P. Venkatapuram; V. Rao; K. Noru; A. Bjorlin; M. Zeile; A. Lewis; A. Singh; A. Sunil; C.-C. Chen; C.-F. Lin; C. Chen; D. P. Sundar; D. Jayaraman; H. Ucar; H. Li; J. Singh; J. C. C. Liu; K. R. Rachamreddy; K. Sriadibhatla; K. Datla; L. V. D. Berg; L. Feng; P. Jampani; R. Moola; R. Mallya; S. Jha; S. Pan; S. Srinivasan; V. Vaduganathan; X. Zha; Z. Wang; A. K. Sengottuvel; B. Alluri; B. Oshin; C. Kanumetta; E. Sahin; J. M. Athaide; J. Wu; K. C. Kurapati; K. Manthati; K. Thottempudi; R. R. Chennamsetti; K. R. Jagannath; S. Arvapalli; T. Kala; T. Wang; P. Chopda; K. Gandhi; A. Ramesh; R. Gupta; S. Fadnavis; A. Qassoud; C. Friedt; F. Li; H. Gao; J. Lee; M. Dixit; S. Ugaji; T. Karuturi; X. Xie; A. Narasimha; B. Jakka; B. Dodds; J. Yang; K. Skandakumaran; M. Modi; P. Modi; C. Stejerean; D. Ronca; H. Wang; N. Pham; L. Lu; H. Shen; J. Ning; K. Narayanan; L. Chen; N. Avidan; W. Arnold; F. Xu; G. Patil; V. Balan; S. D. Grandhi

doi:10.1117/12.2685875

4 October 2023 Efficient video processing at scale using MSVP

H. M. Reddy, Y. Chen, J. Lan, I. Katsavounidis, B. Anandharengan, H. G. Lalgudi, S. Alaparthi, G. Hua, H.-C. Chuang, P.-H. Wu, Z. Lei, A. Mastro, C. Petersen, G. Chaudhari, P. Prakash, S. Regunathan, S. Reddy, P. Venkatapuram, V. Rao, K. Noru, A. Bjorlin, M. Zeile, A. Lewis, A. Singh, A. Sunil, C.-C. Chen, C.-F. Lin, C. Chen, D. P. Sundar, D. Jayaraman, H. Ucar, H. Li, J. Singh, J. C. C. Liu, K. R. Rachamreddy, K. Sriadibhatla, K. Datla, L. V. D. Berg, L. Feng, P. Jampani, R. Moola, R. Mallya, S. Jha, S. Pan, S. Srinivasan, V. Vaduganathan, X. Zha, Z. Wang, A. K. Sengottuvel, B. Alluri, B. Oshin, C. Kanumetta, E. Sahin, J. M. Athaide, J. Wu, K. C. Kurapati, K. Manthati, K. Thottempudi, R. R. Chennamsetti, K. R. Jagannath, S. Arvapalli, T. Kala, T. Wang, P. Chopda, K. Gandhi, A. Ramesh, R. Gupta, S. Fadnavis, A. Qassoud, C. Friedt, F. Li, H. Gao, J. Lee, M. Dixit, S. Ugaji, T. Karuturi, X. Xie, A. Narasimha, B. Jakka, B. Dodds, J. Yang, K. Skandakumaran, M. Modi, P. Modi, C. Stejerean, D. Ronca, H. Wang, N. Pham, L. Lu, H. Shen, J. Ning, K. Narayanan, L. Chen, N. Avidan, W. Arnold, F. Xu, G. Patil, V. Balan, S. D. Grandhi

Author Affiliations +

Proceedings Volume 12674, Applications of Digital Image Processing XLVI; 1267414 (2023) https://doi.org/10.1117/12.2685875
Event: SPIE Optical Engineering + Applications, 2023, San Diego, California, United States

Abstract

Videos uploaded to Meta's Family-of-Apps are transcoded into multiple bitstreams of various codec formats, resolutions and quality to provide the best video quality across the wide variety of devices and connection bandwidth constraints. On Facebook alone, there are more than 4 billion video views per day and to address the video processing at this scale, we needed a video processing solution that can deliver the best video quality possible, with the shortest amount of encoding time — all while being energy efficient, programmable, and scalable. In this paper, we present, Meta Scalable Video Processor (MSVP) that can do video processing at on-par quality compared to SW solutions but at a small fraction of the compute time and energy. Each MSVP ASIC can offer a peak SIMO (Single Input Multiple Output) transcoding performance of 4K at 15fps at the highest quality configuration and can scale up to 4K at 60fps at the standard quality configuration. This performance is achieved at ~10W of PCIe module power. We achieved a throughput gain of ~9x for H.264 when compared against libx264 SW encoding. For VP9, we achieved a throughput gain of ~50x when compared with libVPX speed 2 preset. Key components of MSVP transcoding include video decode, scalar, encoding and quality metric computation. In this paper, we go over ASIC architecture of MSVP, design of individual components and compare the perf/W vs quality against standard industry used SW encoders.

Conference Presentation

(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

H. M. Reddy, Y. Chen, J. Lan, I. Katsavounidis, B. Anandharengan, H. G. Lalgudi, S. Alaparthi, G. Hua, H.-C. Chuang, P.-H. Wu, Z. Lei, A. Mastro, C. Petersen, G. Chaudhari, P. Prakash, S. Regunathan, S. Reddy, P. Venkatapuram, V. Rao, K. Noru, A. Bjorlin, M. Zeile, A. Lewis, A. Singh, A. Sunil, C.-C. Chen, C.-F. Lin, C. Chen, D. P. Sundar, D. Jayaraman, H. Ucar, H. Li, J. Singh, J. C. C. Liu, K. R. Rachamreddy, K. Sriadibhatla, K. Datla, L. V. D. Berg, L. Feng, P. Jampani, R. Moola, R. Mallya, S. Jha, S. Pan, S. Srinivasan, V. Vaduganathan, X. Zha, Z. Wang, A. K. Sengottuvel, B. Alluri, B. Oshin, C. Kanumetta, E. Sahin, J. M. Athaide, J. Wu, K. C. Kurapati, K. Manthati, K. Thottempudi, R. R. Chennamsetti, K. R. Jagannath, S. Arvapalli, T. Kala, T. Wang, P. Chopda, K. Gandhi, A. Ramesh, R. Gupta, S. Fadnavis, A. Qassoud, C. Friedt, F. Li, H. Gao, J. Lee, M. Dixit, S. Ugaji, T. Karuturi, X. Xie, A. Narasimha, B. Jakka, B. Dodds, J. Yang, K. Skandakumaran, M. Modi, P. Modi, C. Stejerean, D. Ronca, H. Wang, N. Pham, L. Lu, H. Shen, J. Ning, K. Narayanan, L. Chen, N. Avidan, W. Arnold, F. Xu, G. Patil, V. Balan, and S. D. Grandhi "Efficient video processing at scale using MSVP", Proc. SPIE 12674, Applications of Digital Image Processing XLVI, 1267414 (4 October 2023); https://doi.org/10.1117/12.2685875

ACCESS THE FULL ARTICLE

PROCEEDINGS
16 PAGES + PRESENTATION

DOWNLOAD PAPER SAVE TO MY LIBRARY

WATCH
PRESENTATION

GET CITATION

KEYWORDS

Video processing

Keywords/Phrases

Search In:

Publication Years