Publication type: Conference paper
Type of review: Peer review (abstract)
Title: Direct data exchange between FPGAs and GPUs using GPUDirect
Authors : Huber, Philipp
Rosenthal, Matthias
et. al : No
Conference details: Embedded World Conference 2020, Nürnberg, 25.-27. Februar 2020
Issue Date: 27-Feb-2020
Language : English
Subjects : FGPA; GPU; GPUDirect; PCIe
Subject (DDC) : 004: Computer science
Abstract: Today’s heterogeneous computer systems combine CPUs, GPUs, and FPGAs with different architectures. GPUs with immense parallelization are best fitted for real-time video and signal processing. On the other hand, FPGAs are ideal for capturing and pre-processing multiple video streams. Therefore, partitioning computational tasks between these computational units and data-paths between them are very important. Close and efficient interaction of the individual components is, therefore, indispensable. With PCIe, modern computers provide a high-speed internal data network between CPU, GPU, and FPGA. However, since PCIe communication usually takes place via the CPU, this communication becomes the bottleneck in applications. This paper compares two different implementations which use PCIe without CPU intervention for an efficient FPGA-GPU co-design. The discussed implementation reaches a transfer jitter of only 4 μs and utilizes 72% of available PCIe bandwidth without any CPU interaction.
Fulltext version : Published version
License (according to publishing contract) : Licence according to publishing contract
Departement: School of Engineering
Organisational Unit: Institute of Embedded Systems (InES)
Appears in Collections:Publikationen School of Engineering

Files in This Item:
There are no files associated with this item.

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.