Effective hardware task context switching in Virtex-4 FPGAs (コンピュータシステム)
スポンサーリンク
概要
- 論文の詳細を見る
A unique aspect of flexibility provided by some of the FPGAs such as Xilinx Virtex-4 family is the capability of dynamic and partial reconfiguration, giving them additional leverage over the other co-existing FPGA solutions by allowing implementation of such concepts as a hardware task. When compared to classical software task effective employment of the new idea in preemptive multitasking systems poses many difficulties and involves many mechanisms such as context saving and restoring, to be built practically from the scratch. This paper presents an effective approach to high-speed context switching for Virtex4-based DPR (Dynamic Partial Reconfiguration) Systems based on developed embedded system infrastructure with lightweight control bus, enhancing management of reconfigurable hardware modules and very efficient, instruction-driven reconfiguration/readback controller which offers 78-fold speed-ups and further superior functionalities when compared to baseline IP provided by FPGA's manufacturer. The whole system is additionally supported by developed bitstream manipulation tool intended for PC (Personal Computer) and used as a back-end program for current DPR design flow.
- 2010-01-19
著者
-
Honda Shinya
Graduate School of Information Science, Nagoya University
-
Tomiyama Hiroyuki
Graduate School of Information Science, Nagoya University
-
Takada Hiroaki
Graduate School of Information Science, Nagoya University
-
Shinya Honda
Graduate School of Information Science, Nagoya University
-
Jozwik Krzysztof
Graduate School of Information Science, Nagoya University
-
Honda Shinya
Graduate School Of Information Science Nagoya University
-
Takada Hiroaki
Graduate School Of Information Science Nagoya University
-
Jozwik Krzysztof
Graduate School Of Information Science Nagoya University
-
Tomiyama Hiroyuki
Graduate School Of Information Science Nagoya University
-
Tomiyama Hiroyuki
The Department Of Computer Science And Communication Engineering Graduate School Of Information Scie
-
Tomiyama Hiroyuki
The Interdisciplinary Graduate School Of Engineering Sciences Kyushu University
-
Takada H
Graduate School Of Information Science Nagoya University
-
Tomiyama H
Department Of Information Engineering The Graduate School Of Information Science Nagoya University
-
Takada H
Department Of Information Engineering The Graduate School Of Information Science Nagoya University
-
Tomiyama H
Graduate School Of Information Science Nagoya University
関連論文
- A realization of RPC in embedded component systems (システムLSI設計技術・組込みシステム・組込技術とネットワークに関するワークショップETNET2008)
- Embedded System Covalidation with RTOS Model and FPGA
- Integrated Scheduling in a Real-Time Embedded Hypervisor
- Effective hardware task context switching in Virtex-4 FPGAs (VLSI設計技術)
- Effective hardware task context switching in Virtex-4 FPGAs (コンピュータシステム)
- Effective hardware task context switching in Virtex-4 FPGAs (リコンフィギャラブルシステム)
- Efficient Design Space Exploration at System Level with Automatic Profiler Instrumentation
- Integrated Scheduling in a Real-Time Embedded Hypervisor
- Instruction Schecduling to Reduce Switching Activity of Off-Chip Buses for Low-Power Systems with Caches (Special Section on VLSI Design and CAD Algorithms)
- Language and Compiler for Optimizing Datapath Widths of Embedded Systems (Special Section on VLSI Design and CAD Algorithms)
- Soft-Core Processor Architecture for Embedded System Design(Special Issue on Novel VLSI Processor Architectures)
- Task migration for energy savings in multiprocessor real-time systems (VLSI設計技術)
- A dynamic algorithm for energy savings in DEPS framework (組込みシステム)
- Power Management for Idle Time in the Presence of Periodic Interrupt Services
- Power Management for Idle Time in the Presence of Periodic Interrupt Services
- Power Management for Idle Time in the Presence of Periodic Interrupt Services
- Power Management for Idle Time in the Presence of Periodic Interrupt Services
- A realization of RPC in embedded component systems (コンピュータシステム・組込技術とネットワークに関するワークショップETNET2008)
- A realization of RPC in embedded component systems (ディペンダブルコンピューティング・組込技術とネットワークに関するワークショップETNET2008)
- Proposal and Quantitative Analysis of the CHStone Benchmark Program Suite for Practical C-based High-level Synthesis
- Memory Data Organization for Low-Energy Address Buses(Low-Power System LSI, IP and Related Technologies)
- Impacts of Compiler Optimizations on Address Bus Energy : An Empirical Study(VLSI Design Technology and CAD)
- Effective Scheduling Algorithms for I/O Blocking with a Multi-Frame Task Model
- An Effective GA-Based Scheduling Algorithm for FlexRay Systems
- A front-end for better behavioral synthesis
- A front-end for better behavioral synthesis
- A Realization of RPC in Embedded Component Systems
- Inter-OS Communications for a Real-Time Dual-OS Monitor
- Inter-OS Communications for a Real-Time Dual-OS Monitor
- HW/SW Cosimulation Framework Based on Software Component System
- HW/SW Cosimulation Framework Based on Software Component System
- Embedded System Cost Optimization via Data Path Width Adjustment (Special Issue on Synthesis and Verification of Hardware Design)
- Satsuki: An Integrated Processor Synthesis and Compiler Generation System (Special Issue on Synthesis and Verification of Hardware Design)
- Partitioning of Behavioral Descriptions with Exploiting Function-Level Parallelism
- Function-Level Partitioning of Sequential Programs for Efficient Behavioral Synthesis
- Function Call Optimization for Efficient Behavioral Synthesis
- An RTOS-Based Design and Validation Methodology for Embedded Systems(System Programs)
- RTOS-Centric Cosimulator for Embedded System Design
- Static Task Scheduling Algorithms Based on Greedy Heuristics for Battery-Powered DVS Systems
- Automatic Communication Synthesis with Hardware Sharing for Multi-Processor SoC Design
- Module Selection Using Manufacturing Information (Special Section on VLSI Design and CAD Algorithms)
- Preemptive Hardware Multitasking on Dynamically Partially Reconfigurable FPGAs : Hardware and Reconfiguration Layers
- A Fast Performance Estimation Framework for System-Level Design Space Exploration (System LSI Design Methodology Vol.5)
- Partitioning and Allocation of Scratch-Pad Memory for Energy Minimization of Priority-Based Preemptive Multi-Task Systems
- A Novel Framework for Effective Preemptive Hardware Multitasking on FPGAs
- Integrated Scheduling for a Reliable Dual-OS Monitor
- Integrated Scheduling for a Reliable Dual-OS Monitor
- Embedded System Covalidation with RTOS Model and FPGA
- Efficient Algorithms for Extracting Pareto-optimal Hardware Configurations in DEPS Framework
- Efficient Design Space Exploration at System Level with Automatic Profiler Instrumentation
- A Fast Performance Estimation Framework for System-Level Design Space Exploration
- Embedded System Covalidation with RTOS Model and FPGA
- Worst Case Response Time Analysis for Messages in Controller Area Network with Gateway
- Proposal and Quantitative Analysis of the CHStone Benchmark Program Suite for Practical C-based High-level Synthesis
- A Channel-based Communication/Synchronization Model for SW-HW Multitasking on Dynamically Partially Reconfigurable FPGAs
- A Channel-based Communication/Synchronization Model for SW-HW Multitasking on Dynamically Partially Reconfigurable FPGAs
- A Channel-based Communication/Synchronization Model for SW-HW Multitasking on Dynamically Partially Reconfigurable FPGAs