1 / 21

CA Final Project – Multithreaded Processor with IPC Interface

CA Final Project – Multithreaded Processor with IPC Interface. Part I – Project Review. Goal. Multicore processors, e.g. the TI OMAP and the IBM Cell, are getting more and more popular either in the multimedia or the communication applications

jasonh
Télécharger la présentation

CA Final Project – Multithreaded Processor with IPC Interface

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. CA Final Project – Multithreaded Processor with IPC Interface Part I – Project Review

  2. Goal • Multicore processors, e.g. the TI OMAP and the IBM Cell, are getting more and more popular either in the multimedia or the communication applications • However, the processors are usually underutilized due to some latency • IPC latency • Pipeline latency • Memory latency • In this project, the studentsare requested to come up with somesolutions to those latency and evaluate the effectiveness

  3. Experiment Environment • ARM Versatile platform • Host processor (ARM926) • DSP accelerator (MIPS compatible) on the FPGA • Baseline (1 thread), 40MHz • IMT (4 threads), 10MHz • Applications • JPEG decoder (VGA-sized)

  4. Grading • Evaluate the DSP utilization in the following cases (the same IPC mechanism) • Multithreaded DSP • MPU handles the IPC (software solution) • DSP handles the IPC (hardware solution) • MPU handles the IPC • Single-threaded DSP • Multithreaded DSP • Explain what is the most critical overhead in your dual-core processor experiment • Explain whether the IMT-DSP is the good solution to reduce the overhead in your experiment

  5. Reference • S. H. Ou, T. J. Lin, X. S. Deng, Z. H. Zhuo, and C. W. Liu, “Multithreaded coprocessor interface for multicore multimedia SoC, ” in ASPDAC, 2008

  6. Introduction to Versatile Programming Part II

  7. Outline • Environment setting • Coding on the CodeWarrier • Burn design on the FPGA

  8. Environment Setting

  9. Software Required • Xilinx ISE (v9.1i) • ARM Multi-ICE (v2.2) • ARM Developer Suite (v1.2)

  10. Xilinx ISE • Install CD1 (main program) • Install CD2 (FPGA model) • Virtex 2 should be included

  11. Multi-ICE • Install Multi-ICE and its patch • The patch contains ARM926EJ-S data • Connect versatile and run “Multi-ICE Server”, 2 mode will be found • Normal mode (config jumper open) • Config mode (config jumper closed)

  12. ARM Developer Suite • Install • Run Multi-ICE first (In normal mode) • Run AXD Server and config as follows: • Option->Configure Target->Add • choose Multi-ICE\Multi-ICE.dll • Configure->select ARM926EJ-S • Load Image and start debug

  13. Coding on theCodeWarrier

  14. Create a “New Project” Include the source code and other files Example

  15. Platform environment settings Select “Make” to generate *axf file and “Run” to run on the ARM926 Example 1. Make Run choose ARM926EJ-S include DMA.scf 2.

  16. Burn Design on the FPGA

  17. AMBA bus connection Address decode AMBA Bus Protocol

  18. RTL codes on Xilinx Project Navigator – Implementation • Translate • Merge multiple design files into a single netlist • Map • Group logical symbols from the netlist (gates) into physical components (slices and IOBs) • Place & Route • Place components onto the chip, connect them, and extract timing data into reports

  19. Wrap your design with a wrapper Generate BIT file Example wrapper your design Click “Generate Programming File”

  20. Multi-Ice mode selection Burn Boardfiles -> progcards_multiice.exe Example

  21. Example • ARM-FPGA co-process

More Related