Novel OpenVX implementation for heterogeneous multi-core systems

Chitnis, Kedar; Villarreal, Jesse; Jadav, Brijesh; Mody, Mihir; Weaver, Lucas; Cheng, Victor; Desappan, Kumar; Jain, Arvind; Swami, Pramod

doi:10.1109/icce-asia.2017.8309323

Cited by 4 publications

(2 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This representation potentially allows the runtime environment to identify opportunities for optimization. Examples of OpenVX vendor implementations are Intel OpenVINO [8], Nvidia VisionWorks [9], AMD OpenVX (AMDOVX) [10], and Texas Instruments TIOVX [11]. In this paper we target the Texas Instrument's TDA2x System-on-Chip and our baseline results are given by the release version of the Texas Instruments Vision SDK with OpenVX framework [12].…”

Section: Openvxmentioning

confidence: 99%

See 1 more Smart Citation

Optimizing OpenVX Graphs for Data Movement

Abeysinghe,

Villarreal,

Bakos

2024

Preprint

View full text Add to dashboard Cite

show abstract

Section: Openvxmentioning

confidence: 99%

“…The combination of image tiling, double buffering with DMA, and scratchpad into the OpenVX runtime on the Texas Instruments C66x DSP provided a benefit of a 2.3 speedup as compared to just using cache [11]. In this case, the runtime decomposes all input and output images of each kernel into 64x48-pixel tiles.…”

Section: Scratchpad Memory and Tilingmentioning

confidence: 99%