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A newer version of this document is available.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 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 High Stall Percentage Low Occupancy Percentage High Stall and High Occupancy Percentages No Stalls, Low Occupancy Percentage, and Low Bandwidth No Stalls, High Occupancy Percentage, and Low Bandwidth High Stall and Low Occupancy Percentages 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

Profiler Analyses of Example SYCL* Design Scenarios

Understanding the problems and solutions presented in example SYCL design scenarios might help you leverage the Profiler metrics of your design to optimize its performance.

High Stall Percentage

A high stall percentage implies that the memory or pipe instruction cannot fulfill the access request because of contention for memory bandwidth or pipe buffer space.

Memory instructions stall often whenever bandwidth usage is inefficient or if a large amount of data transfer is necessary during the execution of your application. Inefficient memory accesses lead to suboptimal bandwidth utilization. In such cases, analyze your kernel memory access for possible improvements.

Pipe instructions stall whenever there is a strong imbalance between read and write accesses to the pipe. Imbalances might be caused by pipe reads or writes operating at different rates.

For example, if you find that the stall percentage of a write pipe call is high, check to see if the occupancy and activity of the read pipe call are low. If they are, the kernel's speed controlling the read pipe call is too slow compared to the kernel controlling the write pipe call, leading to a performance bottleneck.

Low Occupancy Percentage

A low occupancy percentage implies that a work-item is accessing the load and store operations or the pipe infrequently. This behavior is expected for load and store operations or pipes that are in non-critical loops. However, if the memory or pipe instruction is in critical portions of the kernel code and the occupancy or activity percentage is low, it implies that a performance bottleneck exists because work-items or loop iterations are not being issued in the hardware.

Consider the following code example: 

queue_event = deviceQueue.submit([&](handler &cgh) {
   cgh.single_task<class LowOccupancyPercentageEx>([=]() {
      for (int i = 0; i < N; i++) {
         for (int j = 0; j < 1000; j++){
            Pipe1::write(data1);
         }
         for (int k =0; k < 3; k++) {
            Pipe2::write(data2);
         }
      }
   });
});

Assuming all loops are pipelined, the first inner loop with a trip count of 1000 is the critical loop. The second inner loop with a trip count of three is executed infrequently. As a result, you can expect that the occupancy and activity percentages for Pipe1 are high and for Pipe2 are low.

In addition, occupancy percentage might be low if you define a small work-group size, the kernel might not receive sufficient work-items. This is problematic because the pipeline is generally empty for the duration of kernel execution, which leads to poor performance.

High Stall and High Occupancy Percentages

A load and store operation or pipe with a high stall percentage is the cause of the kernel pipeline stall.

NOTE:

An ideal kernel pipeline condition has a stall percentage of 0% and an occupancy percentage of 100%.

Usually, the sum of stall and occupancy percentages approximately equals 100%. If a load and store operation or pipe has a high stall percentage, it means that the load and store operation or pipe has the ability to execute every cycle but is generating stalls.

Solutions for stalling global load and store operations:

  • Use local memory to cache data.
  • Reduce the number of times you read the data.
  • Improve global memory accesses:
    • Change the access pattern for more global-memory-friendly addressing (for example, change from stride accessing to sequential accessing).
    • Compile your kernel with the -Xsno-interleaving=default compiler command option and separate the read and write buffers into different DDR banks.
    • Have fewer but wider global memory accesses.
  • Acquire an accelerator board with more bandwidth (for example, a board with three DDRs instead of two DDRs).

Solution for stalling local load and store operations:

  • Review the Memory Viewer report to verify the local memory configuration and modify the configuration to make it stall-free.

Solutions for stalling pipes:

  • Fix stalls on the other side of the pipe. For example, if a pipe read stalls, it means that the writer to the pipe is not writing data into the pipe fast enough, and you must adjust it.
  • If there are pipe loops in your design, specify the pipe depth.

No Stalls, Low Occupancy Percentage, and Low Bandwidth

Loop-carried dependencies might create a bottleneck in your design that causes an LSU to have a low occupancy percentage and a low bandwidth.

NOTE:

An ideal kernel pipeline has a bandwidth that equals the board's available bandwidth.

Example SYCL Kernel and Profiler Analysis

In this example, dst[] is executed once every 20 iterations of the FACTOR2 loop and once every four iterations of the FACTOR1 loop. Therefore, FACTOR2 loop is the source of the bottleneck.

Solutions for resolving loop bottlenecks:

  • Unroll the FACTOR1 and FACTOR2 loops evenly. Simply unrolling FACTOR1 loop further does not resolve the bottleneck.
  • Vectorize your kernel to allow multiple work-items to execute during each loop iteration.

No Stalls, High Occupancy Percentage, and Low Bandwidth

The structure of a kernel design might prevent it from leveraging all the available bandwidth that the accelerator board can offer.

Example SYCL Kernel and Profiler Analysis

In this example, the accelerator board can provide a bandwidth of 25600 megabytes per second (MB/s). However, the vector_add kernel is requesting (2 reads + 1 write) x 4 bytes x 294 MHz = 12 bytes/cycle x 294 MHz = 3528 GB/s, which is 14% of the available bandwidth. To increase the bandwidth, increase the number of tasks performed in each clock cycle.

Solutions for low bandwidth:

  • Automatically or manually, vectorize the kernel to make wider requests.
  • Unroll the innermost loop to make more requests per clock cycle.
  • Delegate some of the tasks to another kernel.

High Stall and Low Occupancy Percentages

There might be situations where a global store operation might have a high stall percentage (for example, 30%) and very low occupancy percentage (for example, 0.01%). If such a store operation happens once every 10000 cycles of computation, the efficiency of this store is likely not a cause for concern.

Parent topic: Intel® FPGA Dynamic Profiler for DPC++
Level Two Title