Filetype PDF | Posted on 04 Feb 2023 | 2 years ago
Authentication
digital resources
223x Filetype PDF File size 3.21 MB Source: www.intel.com
® ™
Intel FPGA SDK for OpenCL Pro
Edition
Programming Guide
® ®
Updated for Intel Quartus Prime Design Suite: 21.3
Online Version ID: 683846
Send Feedback Version: 2021.10.04
UG-OCL002
Contents
Contents
® ™
1. Intel FPGA SDK for OpenCL Overview......................................................................... 7
1.1. Intel FPGA SDK for OpenCL Pro Edition Programming Guide Prerequisites......................7
1.2. Intel FPGA SDK for OpenCL FPGA Programming Flow..................................................8
2. Intel FPGA SDK for OpenCL Offline Compiler Kernel Compilation Flows........................10
2.1. One-Step Compilation for Simple Kernels................................................................ 11
2.2. Multistep Intel FPGA SDK for OpenCL Pro Edition Design Flow....................................12
3. Obtaining General Information on Software, Compiler, and Custom Platform...............16
3.1. Displaying the Software Version (version)............................................................... 16
3.2. Displaying the Compiler Version (-version).............................................................. 16
3.3. Listing the Intel FPGA SDK for OpenCL Utility Command Options (help).......................17
3.3.1. Displaying Information on an Intel FPGA SDK for OpenCL Utility Command
Option (help )............................................................. 17
3.4. Listing the Intel FPGA SDK for OpenCL Offline Compiler Command Options (no
argument, -help, or -h)......................................................................................17
3.5. Listing the Available FPGA Boards and Custom Platforms (-list-boards and -list-
board-packages)...............................................................................................17
3.6. Displaying the Compilation Environment of an OpenCL Binary (env)........................... 18
4. Managing an FPGA Board..............................................................................................19
4.1. Installing an FPGA Board (install)...........................................................................20
4.2. Uninstalling an FPGA Board (uninstall)....................................................................22
4.3. Querying the Device Name of Your FPGA Board (diagnose)........................................22
4.4. Running a Board Diagnostic Test (diagnose ).....................................23
4.5. Programming the FPGA Offline or without a Host (program )...............23
4.6. Programming the Flash Memory (flash )...........................................23
5. Structuring Your OpenCL Kernel................................................................................... 25
5.1. Guidelines for Naming the Kernel...........................................................................25
5.2. Programming Strategies for Optimizing Data Processing Efficiency..............................26
5.2.1. Unrolling a Loop (unroll Pragma) ............................................................27
5.2.2. Disabling Pipelining of a Loop (disable_loop_pipelining Pragma) .......... 28
5.2.3. Coalescing Nested Loops...........................................................................29
5.2.4. Fusing Adjacent Loops (loop_fuse Pragma)...............................................30
5.2.5. Marking Loops to Prevent Automatic Fusion (nofusion Pragma)................... 32
5.2.6. Specifying a Loop Initiation interval (II)...................................................... 32
5.2.7. Loop Concurrency (max_concurrency Pragma) .........................................34
5.2.8. Loop Speculation (speculated_iterations Pragma) ...............................35
5.2.9. Loop Interleaving Control (max_interleaving Pragma)..............................35
5.2.10. Floating Point Optimizations (fp contract and fp reassociate
Pragma)................................................................................................. 37
5.2.11. Specifying Work-Group Sizes................................................................... 38
5.2.12. Specifying Number of Compute Units........................................................39
5.2.13. Specifying Number of SIMD Work-Items....................................................40
5.2.14. Specifying the private_copies Memory Attribute....................................41
5.2.15. Specifying the use_stall_enable_clusters Cluster-control Attribute ..... 42
5.3. Programming Strategies for Optimizing Pointer-to-Local Memory Size......................... 42
Intel FPGA SDK for OpenCL Pro Edition: Programming Guide
Send Feedback
2
Contents
5.4. Implementing the Intel FPGA SDK for OpenCL Channels Extension............................. 43
5.4.1. Overview of the Intel FPGA SDK for OpenCL Channels Extension.................... 43
5.4.2. Channel Data Behavior............................................................................. 44
5.4.3. Multiple Work-Item Ordering for Channels................................................... 45
5.4.4. Restrictions in the Implementation of Intel FPGA SDK for OpenCL
Channels Extension .................................................................................46
5.4.5. Enabling the Intel FPGA SDK for OpenCL Channels for OpenCL Kernel.............48
5.5. Implementing OpenCL Pipes..................................................................................60
5.5.1. Overview of the OpenCL Pipe Functions.......................................................61
5.5.2. Pipe Data Behavior...................................................................................62
5.5.3. Multiple Work-Item Ordering for Pipes.........................................................62
5.5.4. Restrictions in OpenCL Pipes Implementation...............................................64
5.5.5. Enabling OpenCL Pipes for Kernels............................................................. 65
5.5.6. Direct Communication with Kernels via Host Pipes........................................75
5.6. Implementing Arbitrary Precision Integers...............................................................82
5.7. Using Predefined Preprocessor Macros in Conditional Compilation...............................83
5.8. Declaring __constant Address Space Qualifiers........................................................ 83
5.9. Including Structure Data Types as Arguments in OpenCL Kernels................................84
5.9.1. Matching Data Layouts of Host and Kernel Structure Data Types.....................84
5.9.2. Disabling Insertion of Data Structure Padding ............................................. 86
5.9.3. Specifying the Alignment of a Struct...........................................................86
5.10. Inferring a Register............................................................................................ 87
5.10.1. Inferring a Shift Register.........................................................................88
5.11. Enabling Double Precision Floating-Point Operations................................................89
5.12. Single-Cycle Floating-Point Accumulator for Single Work-Item Kernels.......................89
5.12.1. Programming Strategies for Inferring the Accumulator................................ 90
5.13. Integer Promotion Rules......................................................................................91
6. Designing Your Host Application...................................................................................93
6.1. Host Programming Requirements...........................................................................93
6.1.1. Host Machine Memory Requirements.......................................................... 93
6.1.2. Host Binary Requirement.......................................................................... 93
6.1.3. Multiple Host Threads...............................................................................94
6.1.4. Out-of-order Command Queues................................................................. 94
6.1.5. Requirement for Multiple Command Queues to Execute Kernels Concurrently...94
6.2. Allocating OpenCL Buffers for Manual Partitioning of Global Memory........................... 94
6.2.1. Partitioning Buffers Across Multiple Interfaces of the Same Memory Type.........95
6.2.2. Partitioning Buffers Across Different Memory Types (Heterogeneous Memory).. 96
6.2.3. Creating a Pipe Object in Your Host Application............................................ 98
6.2.4. Enabling All Global Memory....................................................................... 98
6.3. Triggering Collection Profiling Data During Kernel Execution.......................................99
6.3.1. Profiling Autorun Kernels ........................................................................101
6.3.2. Profile Data Acquisition........................................................................... 103
6.4. Accessing Custom Platform-Specific Functions........................................................103
6.5. Modifying Host Program for Structure Parameter Conversion....................................104
6.6. Managing Host Application...................................................................................105
6.6.1. Displaying Example Makefile Fragments (example-makefile or makefile)........105
6.6.2. Compiling and Linking Your Host Application.............................................. 106
6.6.3. Using OpenCL ICD Extension APIs............................................................ 110
6.6.4. Programming an FPGA via the Host.......................................................... 110
6.6.5. Termination of the Runtime Environment and Error Recovery....................... 115
Intel FPGA SDK for OpenCL Pro Edition: Programming Guide
Send Feedback
3
Contents
6.7. Allocating Shared Memory for OpenCL Kernels Targeting SoCs..................................115
6.8. Sharing Multiple Devices Across Multiple Host Programs.......................................... 117
7. Compiling Your OpenCL Kernel....................................................................................119
7.1. Compiling Your Kernel to Create Hardware Configuration File................................... 119
7.2. Compiling Your Kernel without Building Hardware (-c).............................................120
7.3. Compiling and Linking Your Kernels or Object Files without Building Hardware (-rtl)....120
7.4. Specifying the Location of Header Files (-I=).........................................121
7.5. Specifying the Name of an Intel FPGA SDK for OpenCL Offline Compiler Output File
(-o )..............................................................................................122
7.6. Compiling a Kernel for a Specific FPGA Board and Custom Platform (-
board=) and (-board-package=)................122
7.7. Resolving Hardware Generation Fitting Errors during Kernel Compilation (-high-effort) 124
7.8. Specifying Schedule F Target for Kernels (-clock=) .......................125
max
7.9. Defining Preprocessor Macros to Specify Kernel Parameters (-D)........125
7.10. Generating Compilation Progress Report (-v)........................................................126
7.11. Displaying the Estimated Resource Usage Summary On-Screen (-report).................128
7.12. Suppressing Warning Messages from the Intel FPGA SDK for OpenCL Offline
Compiler (-W).................................................................................................128
7.13. Converting Warning Messages from the Intel FPGA SDK for OpenCL Offline
Compiler into Error Messages (-Werror)..............................................................128
7.14. Removing Debug Data from Compiler Reports and Source Code from the .aocx
File (-g0)........................................................................................................129
7.15. Disabling Burst-Interleaving of Global Memory (-no-
interleaving=)...............................................................129
7.16. Forcing Ring Interconnect for Global Memory (-global-ring)....................................130
7.17. Forcing a Single Store Ring to Reduce Area at the Expense of Write Throughput to
Global Memory (-force-single-store-ring)............................................................130
7.18. Forcing Fewer Read Data Reorder Units to Reduce Area at the Expense of Read
Throughput to Global Memory (-num-reorder).....................................................130
7.19. Configuring Constant Memory Cache Size (-const-cache-bytes=)..................... 130
7.20. Relaxing the Order of Floating-Point Operations (-ffp-reassociate)...................131
7.21. Reducing Floating-Point Rounding Operations (-ffp-contract=fast)...........................131
7.22. Speeding Up Your OpenCL Compilation (-fast-compile)..........................................132
7.23. Compiling Your Kernel Incrementally (-incremental)..............................................132
7.23.1. The Incremental Compile Report.............................................................133
7.23.2. Additional Command Options for Incremental Compilation......................... 135
7.23.3. Limitations of the Incremental Compilation Feature...................................137
7.24. Compiling Your Kernel with Memory Error Correction Coding (-ecc)......................... 137
7.25. Disabling Hardware Kernel Invocation Queue (-no-hardware-kernel-invocation-
queue)...........................................................................................................138
7.26. Modifying the Handshaking Protocol (-hyper-optimized-handshaking)......................138
7.27. Pipelining Loops in Non-task Kernels (-auto-pipeline)............................................ 139
8. Emulating and Debugging Your OpenCL Kernel...........................................................140
8.1. Setting up the Emulator......................................................................................140
8.2. Modifying Channels Kernel Code for Emulation....................................................... 141
8.2.1. Emulating a Kernel that Passes Pipes or Channels by Value..........................142
8.2.2. Emulating Channel Depth........................................................................143
8.2.3. Emulating Applications with a Channel That Reads or Writes to an I/O
Channel................................................................................................ 143
8.3. Compiling a Kernel for Emulation (-march=emulator)..............................................144
Intel FPGA SDK for OpenCL Pro Edition: Programming Guide
Send Feedback
4
no reviews yet
Please Login to review.
