• Home
  • Search
  • OpenCL library of stream memory components targeting FPGAs
  • Cite Icon12
  • https://doi.org/10.1109/fpt.2015.7393134Copy DOI Icon

OpenCL library of stream memory components targeting FPGAs

  • Dec 1, 2015
  • Jasmina Vasiljevic +6 more
Show More
  • Abstract
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

In recent years, high-level languages and compilers, such as OpenCL have improved both productivity and FPGA adoption on a wider scale. One of the challenges in the design of high-performance stream FPGA applications is iterative manual optimization of the numerous application buffers (e.g., arrays, FIFOs and scratch-pads). First, to achieve the desired throughput, the programmer faces the burden of analyzing the memory accesses of each application buffer, and based on observed data locality determines the optimal on-chip buffering, and off-chip read/write data access strategy. Second, to minimize throughput bottlenecks, the programmer has to carefully partition the limited on-chip memory resources among many application buffers. In this work we present an FPGA OpenCL library of pre-optimized stream memory components (SMCs). The library contains three types of SMCs, which implement frequently applied data transformations: 1) stencil, 2) transpose and 3) tiling. The library generates SMCs that are optimized both for the specific data transformation they perform as well as the user specified data set size. Further, to ease the partitioning of on-chip memory resources among many application memories, the library automatically maps application buffers to on-chip and off-chip memory resources. This is achieved by enabling the programmer to specify an on-chip memory budget for each component. In terms of on-chip memory, the SMCs perform data buffering to exploit data locality and maximize reuse. In terms of off-chip memory accesses, the SMCs optimize read/write memory operations by performing data coalescing, bursting and prefetching. We show that using the SMC library, the programmer can quickly generate scalable, pre-optimized stream application memory components, thus reaching throughput targets without time consuming manual memory optimization.

Similar Papers
  • Conference Article
  • Citations15

A Memory-Efficient Hashing by Multi-Predicate Bloom Filters for Packet Classification

  • Apr 01, 2008
  • H Yu +1
  • Conference Article
  • Citations3

Boosting Memory Performance of Many-Core FPGA Device through Dynamic Precedence Graph

  • Apr 01, 2013
  • Yu Bai +4
  • Conference Article
  • Citations10

Custom-sized caches in application-specific memory hierarchies

  • Dec 01, 2015
  • Felix Winterstein +4
  • Research Article
  • Citations40

A Resource-Efficient Hardware Architecture for Connected Component Analysis

  • Jul 01, 2016
  • IEEE Transactions on Circuits and Systems for Video Technology
  • Michael J Klaiber +3
  • Research Article
  • Citations6

Real-Time Modeling of Wheel-Rail Contact Laws with System-On-Chip

  • May 01, 2010
  • IEEE Transactions on Parallel and Distributed Systems
  • Yongji Zhou +2
  • Conference Article
  • Citations4

Merging Everything (ME)

  • Jun 02, 2019
  • Xiaoming Chen +3
  • Conference Article
  • Citations5

Memory access optimization for computations on unstructured meshes

  • Aug 01, 2012
  • Antal Hiba +2
  • Conference Article

Proposal of Highly Efficient Memory Access Method Using Locked-Cache on Soft-Core Processor with SIMD Operations

  • Dec 01, 2015
  • Kosuke Hiraishi +3
  • Dissertation
  • Citations4

A novel access pattern-based multi-core memory architecture

  • Dec 18, 2014
  • Tassadaq Hussain
  • Research Article
  • Citations7

A novel hardware-oriented ultra-high-speed object detection algorithm based on convolutional neural network

  • Dec 21, 2019
  • Journal of Real-Time Image Processing
  • Jianquan Li +5
  • Conference Article
  • Citations30

Optimizing SDRAM bandwidth for custom FPGA loop accelerators

  • Feb 22, 2012
  • Samuel Bayliss +1
  • Conference Article
  • Citations43

Unified on-chip memory allocation for SIMT architecture

  • Jun 10, 2014
  • Ari B Hayes +1
  • Research Article
  • Citations4

Optimizing memory access traffic via runtime thread migration for on-chip distributed memory systems

  • Jun 24, 2014
  • The Journal of Supercomputing
  • Weiwei Fu +3
  • Conference Article
  • Citations7

Balancing Memory Accesses for Energy-Efficient Graph Analytics Accelerators

  • Jul 01, 2019
  • Mingyu Yan +10
  • Research Article
  • Citations27

Fast density-based clustering through dataset partition using graphics processing units

  • Oct 19, 2014
  • Information Sciences
  • Woong-Kee Loh +1
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.