Multimedia execution hardware accelerator

Citation
E. Hakkennes et S. Vassiliadis, Multimedia execution hardware accelerator, J VLSI S P, 28(3), 2001, pp. 221-234
Citations number
17
Categorie Soggetti
Eletrical & Eletronics Engineeing
Journal title
JOURNAL OF VLSI SIGNAL PROCESSING SYSTEMS FOR SIGNAL IMAGE AND VIDEO TECHNOLOGY
ISSN journal
13875485 → ACNP
Volume
28
Issue
3
Year of publication
2001
Pages
221 - 234
Database
ISI
SICI code
1387-5485(2001)28:3<221:MEHA>2.0.ZU;2-R
Abstract
In this paper we show that some expressions frequently used in multimedia a pplications can be formulated as a general add-multiply-add operation. We f urther show a hardwired implementation of the Add-Multiply-Add instruction which is no more complex than the multiplier implementation. Furthermore we show that two frequently motion estimation operations, the Sum and Mean of Absolute Differences, can be implemented in hardware requiring also approx imately the same cycle time as the multiplication. We also show that our ap proach can be extended easily to provide the computation of the Sum and Mea n of Absolute Difference of a 16x16 pixel block in no more than four machin e cycles. Additionally we propose a codec hardwired mechanism for the Paeth predictor used in the Portable Network Standard (PNG) that requires at mos t two general purpose ALU cycles. We extend the paeth unit to include the m edian, maximum and minimum operations on three inputs with no additional cy cle time and we also extend the Add-Multiply-Add unit to include the mean o f three numbers. Finally we propose a multimedia hardware accelerator to ac commodate all the proposed operations. The proposed unit is an extension of the multiply pipeline with ALU extensions with no extra stages added. The unit operates on 32 instructions in total.