SDAccel Design Contest: Vivado
1 like279 views
This document provides an overview of the Xilinx Vivado design suite for FPGA hardware design. It discusses the main steps in the hardware design flow including synthesis, place and route, and bitstream generation. It then demonstrates implementing designs for vector addition using both AXI master and AXI stream interfaces. The document walks through creating a Vivado project, adding IP cores like the MicroBlaze processor and memory interface, connecting the design, and generating a bitstream for the FPGA. It also discusses exporting the hardware to Xilinx SDK and running code on the MicroBlaze to test the designs.
SDAccel Design Contest: Vivado
- 1. Courses @ NECST Lorenzo Di Tucci <lorenzo.ditucci@polimi.it> Emanuele Del Sozzo <emanuele.delsozzo@polimi.it> Marco D. Santambrogio <marco.santambrogio@polimi.it> Xilinx Vivado 25/01/2018
- 2. Agenda • Hardware Design Flow (2nd step) • Xilinx Vivado –Synthesis –Place & Route –Bitstream Generation • Hands on example: implementation of both master AXI and AXI stream designs for vector addition using Vivado and SDK
- 3. Reminder! Use this Google Doc to provide your data https://goo.gl/FRCG6y First, install the VPN we have provided you. (Mac: Tunnelblick - Windows/Linux: OpenVPN) To SSH to the machine: ssh <name>.<surname>@nags31.local.necst.it password: user
- 4. Reminder! You can change your password here: http://changepassword.local.necst.it/ You can also RDP to the instance using • Microsoft Remote Desktop (Microsoft/Mac OS) • Remmina (Linux) To connect to the machine, or change your password you must have started the VPN.
- 5. Hardware Design Flow for HPC • Hardware Design Flow (HDF): process to realize a hardware module • HDF for FPGAs can be seen as a 2 step process
- 6. The Hardware Design Flow
- 7. The Hardware Design Flow System integration, driver generation and runtime management
- 8. The Hardware Design Flow System integration, driver generation and runtime management
- 9. The Hardware Design Flow System integration, driver generation and runtime management
- 10. Vivado Design Suite • Vivado Design Suite is a software suite for synthesis and analysis of HDL designs • Vivado enables developers to synthesize designs, perform timing analysis, examine RTL diagrams, simulate designs, and configure the target FPGA • Starting from HDL, Vivado performs several steps to eventually generate the bitstream
- 11. Vivado Main Steps • Synthesis: translation from HDL to gate level • Place: placing of all the logic components on the FPGA • Route: design of all the wires needed to connect the placed components • Bitstream: generation of FPGA configuration file
- 12. Synthesis
- 13. Place & Route
- 14. Bitstream
- 15. Launch Vivado Source settings64.sh file and launch vivado
- 25. Project Window • text Run / Generate Bitstream / Settings / Project Summary
- 26. Create Block Design • text
- 27. Block Design • text Add IP Validate Design Regenerate Layout Tcl Console Settings
- 28. Add IP • text
- 29. Add Memory Interface Generator • text
- 30. Run Block Automation • text
- 31. Block Automation Settings • text
- 32. Run Connection Automation • text
- 33. Connection Automation Settings • text
- 34. MIG Block • text
- 35. Customize MIG Block • text
- 42. MIG Customization • text This page allows to add / customize clock signals MIG will generate two clocks: - ui_clk (200Mhz) - ui_addn_clk_0 (100Mhz)
- 55. Add IP • text
- 57. Run Block Automation • text
- 58. Block Automation Settings • text
- 59. Run Connection Automation • text
- 60. Connection Automation Settings • text
- 61. Run Connection Automation • text
- 62. Connection Automation Settings • text
- 63. MIG + Microblaze Design • text
- 65. Settings • text
- 67. Add IP Repository • text
- 68. Path to IP Repository • text
- 69. IP Found • text
- 71. Add IP • text
- 73. Connect Kernel Clock • text Connect Kernel Clock to Microblaze clock
- 74. Connection Automation Settings • text
- 76. Enable JTAG UART • text
- 77. Run Connection Automation • text
- 78. Connection Automation Settings • text
- 79. Add IP • text
- 80. Add AXI Timer • text
- 81. Run Connection Automation • text
- 82. Connection Automation Settings • text
- 87. Sources • text
- 89. HDL Wrapper Creation • text
- 91. Launch Synthesis And Implementation • text
- 93. Bitstream Generation Completed • text
- 97. Export Hardware Settings • text This command will create an HDF archive containing the bitstream and drivers for SDK
- 99. Launch SDK Settings • text
- 100. SDK Window • text
- 102. New Project • text
- 103. Templates • text
- 104. Program FPGA • text
- 105. Program FPGA settings • text
- 109. Target Setup • text
- 111. Target Setup done • text
- 112. Application • text
- 116. STDIO Connection done • text
- 117. Hello World Execution • text
- 118. Lscript • text
- 119. Running Our Design • text
- 120. AXI Stream Design • Let’s know implement the design for the streaming version of the vector addition • For this design, we will use the DMA IP • All the steps until the “MIG + Microblaze Design” slide do not change
- 121. IP Settings • text
- 122. Settings • text
- 123. Repository • text
- 124. Add IP Repository • text
- 125. Path to IP Repository • text
- 126. IP Found • text
- 128. Add IP • text
- 129. Add AXI DMA (0) • text
- 130. Customize DMA (0) • text
- 131. DMA Settings (0) • text
- 132. Connect DMA (0) clocks • text
- 133. Connection Automation Settings • text
- 134. Add IP • text
- 135. Add Kernel • text
- 136. Connect Kernel clock • text
- 137. Connect Kernel Reset • text
- 138. Connect port b to DMA (0) • text
- 139. Connect port a to DMA (0) • text
- 140. Add IP • text
- 141. Add AXI DMA (1) • text
- 142. Customize DMA (1) • text
- 143. DMA Settings (1) • text
- 144. Connect DMA (1) clocks • text
- 145. Connection Automation Settings • text
- 146. Connect port c to DMA (1) • text
- 147. Add IP • text
- 148. Add AXI DMA (2) • text
- 149. Customize DMA (2) • text
- 150. DMA (2) Settings • text
- 151. Connect DMA (2) clocks • text
- 152. Connection Automation Settings • text
- 153. Connect port d to DMA (2) • text
- 154. Next steps It is now possible to repeat the steps done for the previous design: –enable JTAG UART on the MDM block –add AXI Timer IP –Run Validation –Create HDL Wrapper –Run Generate Bitstream –Export Hardware –Run SDK
- 157. SDK steps After exporting the hardware from Vivado: – Launch SDK – Create the project – Test the “helloworld” – Use the provided code to evaluate the design
- 158. Summary • Vivado toolchain allows developers to design both the IP they want to accelerate and the overall system • Vivado/Vivado HLS examples, and SDK code are available on nags31 server in /sdaccel_contest folder • Next lectures will focus on SDAccel toolchain
- 159. Feedbacks • We are working at improving this course, would you share your feedback for this lesson? https://goo.gl/tLcWQj
- 160. Thank You for the Attention! Lorenzo Di Tucci lorenzo.ditucci@polimi.it Emanuele Del Sozzo emanuele.delsozzo@polimi.it Marco D. Santambrogio marco.santambrogio@polimi.it