Optimizing inter-FPGA communication by automatic channel adaptation

Romoth, Johannes; Jungewelter, Dirk; Hagemeyer, Jens; Porrmann, Mario; Rückert, Ulrich

doi:10.1109/reconfig.2012.6416767

Cited by 3 publications

(2 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Because of the advantage of customizability and high speed of FPGA's, they are selected over DSP kits [10]. Due to their high embedded memories, FPGA's are used in communication protocols [14], replacing microcontrollers which were used in robotics [15], used in control application of inverter operation [16] as well as used in power electronic converters [17]. The paper compares the results obtained after the synthesis of the proposed FFT based architecture with the DHT based architecture.…”

Section: Daniel Rajumentioning

confidence: 99%

Efficient architecture for real time implementation of Hilbert Transform in FPGA

Prince

Verma

Jayakumar

et al. 2015

2015 IEEE International Conference on Electrical, Computer and Communication Technologies (ICECCT)

View full text Add to dashboard Cite

This paper presents an architecture for real time hardware implementation of Hilbert Transform (HT) using Fast Fourier Transform (FFT). HT is studied and its various application areas are discussed in the paper. Two different architectures are proposed using Fast Fourier Transform (FFT) for the implementation. Implementation of HT using the proposed FFT based architectures are compared with the implementations using Discrete Fourier Transform (DFT) and Discrete Hartley Transform (DHT). The proposed FFT based architectures are implemented on Xilinx Kintex-7 based FPGA and the results acquired are presented in comparison to results obtained through MATLAB simulations. The architecture implemented supports transform length of 8192 points as a demonstrator to the idea using 24 bit fixed point arithmetic.Detailed comparison study in terms of resource utilization and timing analysis is also carried out and the results are reported.

show abstract

Section: Daniel Rajumentioning

confidence: 99%

Efficient architecture for real time implementation of Hilbert Transform in FPGA

Prince

Verma

Jayakumar

et al. 2015

2015 IEEE International Conference on Electrical, Computer and Communication Technologies (ICECCT)

View full text Add to dashboard Cite

show abstract

“…Moreover, FPGAs are well-known for their high parallelization that allows them to perform multiple tasks simultaneously with low latency processing times. This capability makes them suitable for real-time applications, where time-sensitive tasks request fast and efficient computation [2]. In particular for ML inference tasks, where latency, parallelism and energy efficiency play a crucial role [3].…”

Section: Introductionmentioning

confidence: 99%

KServe inference extension for an FPGA vendor-free ecosystem

Ciangottini,

Bianchini,

Mariotti

et al. 2024

EPJ Web of Conf.

View full text Add to dashboard Cite

Field Programmable Gate Arrays (FPGAs) are playing an increasingly important role in the sampling and data processing industry due to their intrinsically highly parallel architecture, low power consumption, and flexibility to execute custom algorithms. In particular, the use of FPGAs to perform Machine Learning (ML) inference is increasingly growing thanks to the development of High-Level Synthesis (HLS) projects that abstract the complexity of Hardware Description Language (HDL) programming. In this work we will describe our experience extending KServe predictors, an emerging standard for ML model inference as a service on kubernetes. This project will support a custom workflow capable of loading and serving models on-demand on top of FPGAs. A key aspect of the proposed approach is to make the firmware generation, often an obstacle to a widespread FPGA adoption, transparent. We will detail how the proposed system automates both the synthesis of the HDL code and the generation of the firmware, starting from a high-level language and user-friendly machine learning libraries. The ecosystem is then completed with the adoption of a common language for sharing user models and firmwares, that is based on a dedicated Open Container Initiative artifact definition, thus leveraging all the well established practices on managing resources on a container registry.

show abstract