Developer Guide

FPGA Optimization Guide for Intel® oneAPI Toolkits

Download PDF
ID 767853
Date 7/13/2023
Public

A newer version of this document is available. Customers should click here to go to the newest version.

Document Table of Contents
Document Table of Contents x
FPGA Optimization Guide for Intel® oneAPI Toolkits Introduction To FPGA Design Concepts Analyze Your Design Optimize Your Design FPGA Optimization Flags, Attributes, Pragmas, and Extensions Quick Reference Additional Information Document Revision History for the FPGA Optimization Guide for Intel® oneAPI Toolkits
Introduction To FPGA Design Concepts x
FPGA Architecture Overview Concepts of FPGA Hardware Design Methods of Hardware Design How Source Code Becomes a Custom Hardware Datapath Scheduling Mapping Parallelism Models to FPGA Hardware Memory Types
FPGA Architecture Overview x
Adaptive Logic Module (ALM) Lookup Table (LUT) Register Digital Signal Processing (DSP) Block Random Access Memory (RAM) Blocks
Concepts of FPGA Hardware Design x
Maximum Frequency (fMAX) Latency Pipelining Throughput Datapath Control Path Occupancy
How Source Code Becomes a Custom Hardware Datapath x
Mapping Source Code Instructions to Hardware Mapping Arrays and Their Accesses to Hardware
Scheduling x
Dynamic Scheduling Clustering the Datapath Handshaking Between Clusters
Mapping Parallelism Models to FPGA Hardware x
Data Parallelism Task Parallelism
Data Parallelism x
Executing Independent Operations Simultaneously Pipelining
Memory Types x
Kernel Memory Global Memory
Analyze Your Design x
Analyze the FPGA Early Image Analyze the FPGA Image
Analyze the FPGA Early Image x
Review the FPGA Optimization Report Access HLD FPGA Reports in JSON Format
Review the FPGA Optimization Report x
Loop Analysis Loop Pragma and Attributes Key Performance Metrics Example Bottlenecks Viewer Area Estimates System Viewer Kernel Memory Viewer Schedule Viewer
Analyze the FPGA Image x
Quartus (Static) Summary Intel® FPGA Dynamic Profiler for DPC++ System-level Profiling Using the Intercept Layer for OpenCL* Applications
Quartus (Static) Summary x
Timing Failures
Intel® FPGA Dynamic Profiler for DPC++ x
Measure Kernel Performance Instrument the Kernel Pipeline with Performance Counters (-Xsprofile) Obtain Profiling Data During Runtime Reduce Area Resource Use While Profiling Profiler Analyses of Example SYCL* Design Scenarios Limitations
Obtain Profiling Data During Runtime x
Invoke the Profiler Runtime Wrapper to Obtain Profiling Data Use Intel® VTune™ Profiler
Use Intel® VTune™ Profiler x
Interpret Performance Counter Data
System-level Profiling Using the Intercept Layer for OpenCL* Applications x
Set Up the Intercept Layer for OpenCL* Applications
Optimize Your Design x
Throughput Resource Use
Throughput x
Single Work-item Kernels NDRange Kernels Memory Accesses Pipes Host
Single Work-item Kernels x
Single Work-item Kernel Design Guidelines Loops Single-Cycle Floating-Point Accumulator for Single Work-Item Kernels
Loops x
Refactor the Loop-Carried Data Dependency Relax Loop-Carried Dependency Transfer Loop-Carried Dependency to Local Memory Minimize the Memory Dependencies for Loop Pipelining Unroll Loops Fuse Loops to Reduce Overhead and Improve Performance Optimize Loops With Loop Speculation Remove Loop Bottlenecks Shannonization to Improve FMAX/II Optimize Inner Loop Throughput Improve Loop Performance by Caching On-Chip Memory
Single-Cycle Floating-Point Accumulator for Single Work-Item Kernels x
Strategies for Inferring the Accumulator
Memory Accesses x
Load-Store Units Global Memory Accesses Optimization Perform Kernel Computations Using Local or Private Memory Local and Private Memory Accesses Optimization Annotating Unified Shared Memory Pointers Zero-Copy Memory Access Additional Recommendations
Load-Store Units x
Load-Store Unit Styles Load-Store Unit Modifiers Load-Store Unit Controls
Global Memory Accesses Optimization x
Global Memory Bandwidth Use Calculation Manual Partition of Global Memory Partitioning Buffers Across Different Memory Types (Heterogeneous Memory) Partitioning Buffers Across Memory Channels of the Same Memory Type Ignoring Dependencies Between Accessor Arguments Contiguous Memory Accesses Static Memory Coalescing
Pipes x
Host Pipes
Host Pipes x
Host Pipe Declaration Host Pipe API Host Pipes RTL Interfaces
Host x
Multi-Threaded Host Application Utilizing Hardware Kernel Invocation Queue Double Buffering Host Utilizing Kernel Invocation Queue N-Way Buffering to Overlap Kernel Execution Prepinning Memory Simple Host-Device Streaming Buffered Host-Device Streaming
Double Buffering Host Utilizing Kernel Invocation Queue x
Applying Double-Buffering Using the Intercept Layer for OpenCL* Applications
Resource Use x
Data Types and Operations Kernel Variable Accesses
Data Types and Operations x
Optimize Floating-point Operation Avoid Expensive Functions Variable-Precision Integer and Floating-Point Support
Variable-Precision Integer and Floating-Point Support x
Advantages and Limitations of Arbitrary Precision Data Types Declare and Use the AC Data Types
Declare and Use the AC Data Types x
Declare the ac_int Data Type Declare the ac_fixed Data Type Declare the ac_complex Data Type Declare the ap_float Data Type
Declare the ap_float Data Type x
Conversion Rules for ap_float Operations with Explicit Precision Controls Comparison Operators Additional ap_float Functions Additional Data Types Provided by the ap_float.hpp Header File Quality of Results and the ap_float Data Type
FPGA Optimization Flags, Attributes, Pragmas, and Extensions x
Optimization Flags Optimization Targets Kernel Variables Kernel Attributes Memory Attributes Loop Directives Floating-Point Pragmas Latency Controls (Beta) FPGA Extensions
Optimization Flags x
Specify Schedule FMAX Target for Kernels (-Xsclock=<clock target>) Create a 2xclock Interface (-Xsuse-2xclock) Disable Burst-Interleaving of Global Memory (-Xsno-interleaving=<global_memory_name>) Force Ring Interconnect for Global Memory (-Xsglobal-ring) Force a Single Store Ring to Reduce Area (-Xsforce-single-store-ring) Force Fewer Read Data Reorder Units to Reduce Area (-Xsnum-reorder) Disable Hardware Kernel Invocation Queue (-Xsno-hardware-kernel-invocation-queue) Modify the Handshaking Protocol Between Clusters (-Xshyper-optimized-handshaking) Disable Automatic Fusion of Loops (-Xsdisable-auto-loop-fusion) Fuse Adjacent Loops With Unequal Trip Counts (-Xsenable-unequal-tc-fusion) Pipeline Loops in Non-task Kernels (-Xsauto-pipeline) Control Semantics of Floating-Point Operations (-fp-model=<value>) Modify the Rounding Mode of Floating-point Operations (-Xsrounding=<rounding_type>) Global Control of Exit FIFO Latency of Stall-free Clusters (-Xssfc-exit-fifo-type=<value>) Enable the Read-Only Cache for Read-Only Accessors (-Xsread-only-cache-size=<N>) Control Hardware Implementation of the Supported Data Types and Math Operations (-Xsdsp-mode=<option>) Generate Register Map Wrapper (-Xsregister-map-wrapper-type)
Optimization Targets x
Minimum Latency Flow Maximum Throughput Without Area Optimization Heuristics Flow
Kernel Attributes x
Specify Schedule FMAX Target for Kernels Specify a Workgroup Size Specify Number of SIMD Work Items Omit Hardware that Generates and Dispatches Kernel IDs Omit Hardware to Support the no_global_work_offset Attribute in parallel_for Kernels Reduce Kernel Area and Latency
Loop Directives x
disable_loop_pipelining Attribute initiation_interval Attribute ivdep Attribute loop_coalesce Attribute max_concurrency Attribute max_interleaving Attribute speculated_iterations Attribute unroll Pragma Loop Fuse Functions and nofusion Attribute max_reinvocation_delay Attribute (Beta)
FPGA Extensions x
Pipes Extension Asynchronous Parallelism Within Kernels (task_sequence) device_global Extension (Beta)
Pipes Extension x
Key Properties of a Pipe Accessing Pipes The pipe Class and its Use I/O Pipes Characteristics of Pipes Restrictions of Pipes Guidelines for Designing Pipes Pipe and Atomic Fence
Asynchronous Parallelism Within Kernels (task_sequence) x
Task Functions task_sequence Use Cases
Quick Reference x
Algorithmic C Data Types Floating Point Pragmas FPGA Accessor Properties FPGA Extensions FPGA Kernel Attributes FPGA Local Memory Function Latency Control Properties (Beta) FPGA LSU Controls FPGA Loop Directives FPGA Memory Attributes FPGA Optimization Flags Pipe API
Document Revision History for the FPGA Optimization Guide for Intel® oneAPI Toolkits x
Notices and Disclaimers
FPGA Optimization Guide for Intel® oneAPI Toolkits
Introduction To FPGA Design Concepts
Analyze Your Design
Optimize Your Design
FPGA Optimization Flags, Attributes, Pragmas, and Extensions
Quick Reference
Additional Information
Document Revision History for the FPGA Optimization Guide for Intel® oneAPI Toolkits

Loop Analysis

The report.html file contains information about all loops in your design and their unroll statuses. The Loop Analysis report helps you examine whether the Intel® oneAPI DPC++/C++ Compiler can maximize your kernel's throughput.

To view the Loop Analysis report, click Throughput Analysis > Loop Analysis. The purpose of this view is to show estimates of performance indicators (such as II) and potential performance bottlenecks. For each loop, you can identify the following using the report:

  • Whether the loop is pipelined
  • Whether the loop uses a hyper-optimized loop structure
  • Any pragma or attribute applied to the loop
  • II of the loop
NOTE:
  • Loop Analysis report does not report anything about loops in NDRange kernels.
  • FPGA optimization reports support user-defined loop labels and replace the system-generated loop labels.

The left-hand Loops List pane of the Loop Analysis report displays the following types of loops the compiler detects in your design and indicates any transformations the compiler may have applied. These transformations may be applied automatically or due to a source-code annotation, such as a pragma or an attribute.

Key Performance Metrics

The Loop Analysis report captures the following key performance metrics on all blocks:

  • Source Location: Indicates the loop location in the source code.
  • Pipelined: Indicates whether a loop is pipelined. Pipelining allows for many data items to be processed concurrently (in the same clock cycle) while efficiently using of the hardware in the datapath by keeping it occupied.
  • II: Shows the sustainable initiation interval (II) of the loop. Processing data in loops is an additional source of pipeline parallelism. When you pipeline a loop, the next iteration of the loop begins before previous iterations complete. You can determine the number of clock cycles between iterations by the number of clock cycles you require to resolve any dependencies between iterations. You can refer to this number as the initiation interval (II) of the loop. The Intel® oneAPI DPC++/C++ Compiler automatically identifies these dependencies and builds hardware to resolve these dependencies while minimizing the II.
  • Estimated fMAX: Shows the estimated maximum clock frequency at which the loop operates. You can also reference it as "Scheduled fMAX." The fMAX is the maximum rate at which the outputs of registers are updated. If the estimated fMAX is below the target frequency, then the estimated fMAX appears in red color and a question mark with a tooltip displaying the target frequency displays.

    The physical propagation delay of the signal between two consecutive registers limits the clock speed. This propagation delay is a function of the complexity of the Boolean logic in the path. The path with the most logic (and the highest delay) limits the speed of the entire circuit, and you can refer to this path as the critical path.

    The fMAX is calculated as the inverse of the critical path delay. High fMAX is desirable because it correlates directly with high performance in the absence of other bottlenecks. The compiler attempts to optimize for different objectives for the estimated fMAX depending on whether the fMAX target is set and whether the #pragma II is set for each of the loops. The fMAX target is a strong suggestion, and the compiler does not error out if it cannot achieve this fMAX, whereas the #pragma II triggers an error if the compiler is not able to achieve the requested II. The fMAX achieved for each block of code is shown in the Loop Analysis report. This behavior is outlined in the following table:

    Explicitly specify fMAX?

    Explicitly specify II?

    Compiler's Behavior

    No

    No

    Use heuristic to achieve best fMAX/II trade-off.

    No

    Yes

    Best effort to achieve the II for the corresponding loop (may not achieve the best possible fMAX).

    Yes

    No

    Best effort to achieve fMAX specified (may not achieve the best possible II).

    Yes

    Yes

    Best effort to achieve the fMAX specified at the given II. The compiler errors out if it cannot achieve the requested II.

    NOTE:

    Intel® recommends that if you are using an fMAX target in the command line or for a kernel, use #pragma II = <N> for performance-critical loops in your design.

  • Latency: Shows the number of clock cycles a loop takes to complete one or more instructions. Typically, you want to have low latency. However, lowering latency often results in decreased fMAX.
  • Speculated Iterations: Shows the loop speculation. Loop speculation is an optimization technique that enables more efficient loop pipelining by allowing future iterations to be initiated before determining whether the loop was exited already.
  • Max Interleaving Iterations: Indicates the number of interleaved invocations of an inner loop that can be executed simultaneously. For more information, refer to max_interleaving Attribute.

Example

The following is a SYCL* kernel example that includes three loops: 

 cgh.single_task<class example>([=]() {
   #pragma unroll
   for (int i = 0; i < 10; i++) {
     acc_data[i] += i;
   }
   #pragma unroll 1
   for (int k = 0; k < N; k++) {
     #pragma unroll 5
     for (int j = 0; j < N; j++) {
       acc_data[j] = j + k;
     }
   }
 });

The Loop Analysis report of this design example highlights the unrolling strategy for the different kinds of loops defined in the code.

The Intel® oneAPI DPC++/C++ Compiler implements the following loop unrolling strategies based on the source code:

  • Fully unrolls the first inner loop (line 3) because of the #pragma unroll specification.
  • Does not unroll the second loop (line 7), which is an outer loop because of the #pragma unroll 1 specification.
  • Unrolls the third loop (line 9, an inner loop of the second loop) five times because of the #pragma unroll 5 specification.

For more examples, refer to Loops section.

Parent topic: Review the FPGA Optimization Report
Level Two Title