Show simple item record

dc.contributor.authorTikekar, Mehul
dc.contributor.authorSze, Vivienne
dc.contributor.authorChandrakasan, Anantha P
dc.date.accessioned2021-10-27T20:10:16Z
dc.date.available2021-10-27T20:10:16Z
dc.date.issued2018
dc.identifier.urihttps://hdl.handle.net/1721.1/135005
dc.description.abstract© 1966-2012 IEEE. This paper proposes a fully integrated H.265/high efficiency video coding (HEVC) video decoder that supports real-time video playback within the 50-mW power budget of wearable devices, such as smart watches and virtual reality (VR) headsets. Specifically, this paper focuses on reducing data movement to and from off-chip memory as it dominates energy consumption, consuming 2.8-6 times more energy than processing in most video decoders. Embedded dynamic random access memory (eDRAM) is used for main memory, and several techniques are proposed to reduce the power consumption of the eDRAM itself: 1) lossless compression is used to store reference frames in two times fewer eDRAM macros, reducing refresh power by 33%; 2) eDRAM macros are powered up on-demand to further reduce refresh power by 33%; and 3) syntax elements are distributed to four decoder cores in a partially compressed form to reduce decoupling buffer power by four times. These approaches reduce eDRAM power by two times in a fully integrated H.265/HEVC decoder with the lowest reported system power. The test chip containing 10.5 MB of eDRAM requires no external memory and consumes 24.9-30.6 mW for decoding 1920× 1080 video at 24-50 frames/s.
dc.language.isoen
dc.publisherInstitute of Electrical and Electronics Engineers (IEEE)
dc.relation.isversionof10.1109/JSSC.2018.2837124
dc.rightsCreative Commons Attribution-Noncommercial-Share Alike
dc.rights.urihttp://creativecommons.org/licenses/by-nc-sa/4.0/
dc.sourceOther repository
dc.titleA Fully Integrated Energy-Efficient H.265/HEVC Decoder With eDRAM for Wearable Devices
dc.typeArticle
dc.contributor.departmentMassachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
dc.relation.journalIEEE Journal of Solid-State Circuits
dc.eprint.versionAuthor's final manuscript
dc.type.urihttp://purl.org/eprint/type/JournalArticle
eprint.statushttp://purl.org/eprint/status/PeerReviewed
dc.date.updated2019-05-22T16:05:15Z
dspace.orderedauthorsTikekar, M; Sze, V; Chandrakasan, AP
dspace.date.submission2019-05-22T16:05:16Z
mit.journal.volume53
mit.journal.issue8
mit.metadata.statusAuthority Work and Publication Information Needed


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record