Developer Guide

FPGA Optimization Guide for Intel® oneAPI Toolkits

Download PDF
ID 767853
Date 12/16/2022
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A newer version of this document is available.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
FPGA Optimization Guide for Intel® oneAPI Toolkits x
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 Notices and Disclaimers
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 System View Global Memory View Block View Cluster View 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 (<span class='codeph'>-Xsprofile</span>) 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
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 <span class='codeph'>ac_int</span> Data Type Declare the <span class='codeph'>ac_fixed</span> Data Type Declare the <span class='codeph'>ac_complex</span> Data Type Declare the <span class='codeph'>ap_float</span> Data Type
Declare the <span class='codeph'>ap_float</span> Data Type x
Conversion Rules for <span class='codeph'>ap_float</span> Operations with Explicit Precision Controls Comparison Operators Additional <span class='codeph'>ap_float</span> Functions Additional Data Types Provided by the <span class='codeph'>ap_float.hpp</span> Header File Quality of Results and the ap_float Data Type
FPGA Optimization Flags, Attributes, Pragmas, and Extensions x
Optimization Flags Kernel Attributes Kernel Controls Kernel Variables Memory Attributes Loop Directives Floating-Point Pragmas Latency Controls (Beta) System of Tasks Extension (<span class='codeph'>task_sequence</span>)
Optimization Flags x
Specify Schedule FMAX Target for Kernels (<span class='codeph'>-Xsclock=<clock target>) Disable Burst-Interleaving of Global Memory (<span class='codeph'>-Xsno-interleaving=<global_memory_type></span>) Force Ring Interconnect for Global Memory (<span class='codeph'>-Xsglobal-ring</span>) Force a Single Store Ring to Reduce Area (<span class='codeph'>-Xsforce-single-store-ring</span>) Force Fewer Read Data Reorder Units to Reduce Area (<span class='codeph'>-Xsnum-reorder</span>) Disable Hardware Kernel Invocation Queue (<span class='codeph'>-Xsno-hardware-kernel-invocation-queue</span>) Modify the Handshaking Protocol Between Clusters (<span class='codeph'>-Xshyper-optimized-handshaking</span>) Disable Automatic Fusion of Loops (<span class='codeph'>-Xsdisable-auto-loop-fusion</span>) Fuse Adjacent Loops With Unequal Trip Counts (<span class='codeph'>-Xsenable-unequal-tc-fusion</span>) Pipeline Loops in Non-task Kernels (<span class='codeph'>-Xsauto-pipeline</span>) Control Semantics of Floating-Point Operations (<span class='codeph'>-fp-model=<var><value></var> </span>) Modify the Rounding Mode of Floating-point Operations (<span class='codeph'>-Xsrounding=<rounding_type></span>) Global Control of Exit FIFO Latency of Stall-free Clusters (<span class='codeph'>-Xssfc-exit-fifo-type=<var><value></var> </span>) Enable the Read-Only Cache for Read-Only Accessors (<span class='codeph'>-Xsread-only-cache-size=<var><N></var>)</span> Control Hardware Implementation of the Supported Data Types and Math Operations (<span class='codeph'>-Xsdsp-mode=<var><option></var> </span>)
Kernel Attributes x
Specify Schedule FMAX Target for Kernels Specify a Workgroup Size Specify Number of SIMD WorkItems Omit Hardware that Generates and Dispatches Kernel IDs Omit Hardware to Support the <span class='codeph'>no_global_work_offset</span> Attribute in <span class='codeph'>parallel_for</span> Kernels Reduce Kernel Area and Latency
Kernel Controls x
Pipes Extension
Pipes Extension x
Key Properties of a Pipe Accessing Pipes The <span class='codeph'>pipe</span> Class and its Use I/O Pipes Characteristics of Pipes Restrictions of Pipes Guidelines for Designing Pipes Pipe and Atomic Fence
Loop Directives x
<span class='codeph'>disable_loop_pipelining</span> Attribute <span class='codeph'>initiation_interval</span> Attribute <span class='codeph'>ivdep</span> Attribute <span class='codeph'>loop_coalesce</span> Attribute <span class='codeph'>max_concurrency</span> Attribute <span class='codeph'>max_interleaving</span> Attribute <span class='codeph'>speculated_iterations</span> Attribute <span class='codeph'>unroll</span> Pragma Loop Fuse Functions and <span class='codeph'>nofusion</span> Attribute
System of Tasks Extension (<span class='codeph'>task_sequence</span>) x
Task Functions <span class='codeph'>task_sequence</span> 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 <span class='codeph'>task_sequence</span> Template Parameters and Function APIs
FPGA Optimization Guide for Intel® oneAPI Toolkits

System Viewer

The <project_dir>/reports/report.html file provides a stall-point graph that includes sizes and types of loads and stores, stalls, latencies, load and store information between kernels and different memories, pipes connected between kernels, and loops.

The System Viewer (see Figure 1) shows an abstracted netlist of your DPC++ system in a hierarchical graphical report consisting of system, global memory, block, and cluster views. It allows you to review information such as sizes, types, dependencies, and schedules of instructions, FIFO created by feedback nodes, FIFO created for capacity balancing in stallable regions, properties of interfaces such as pipe and memory interface, and view variables with loop-carried dependencies.

To view the System Viewer, click Views > System Viewer.

You can interact with the System Viewer in the following ways:

  • Use the mouse wheel to zoom in and out within the System Viewer.
  • Navigate through the following hierarchical views in the left-hand pane:
    • System
    • Global memory
    • Block
    • Cluster
  • Click a node to display its location in the source code in the Code pane and node details in the Details pane.

System View

Use the system view of the System Viewer report to view various kernels in your system. The system view illustrates connections between your kernels and connections from kernels to memories. In addition, the system view shows the connection of blocks within a kernel and highlights blocks with a high initiation interval (II).

Kernel System View of the System Viewer Report

Global Memory View

The global memory view of the System Viewer provides a list of all global memories in the design. The global memory view shows the following:

  • Connectivity within the system showing data flow direction between global memory and kernels.
  • Memory throughput bottlenecks.
  • Status of the compiler flags, such as -Xsnum-reorder and -Xsforce-single-store-ring.
  • Global load-store unit (LSU) types.
  • Type of write/read interconnects.
  • Number of write rings.
  • Number and connectivity of read-router buses.

The following image is an example of the global memory view in the System Viewer:

Graphical Representation of the Global Memory in the System Viewer

In Figure 2:

  • When you select stores or loads, you can view respective lines in the source code and details about the LSU type and LSU-level bandwidth.
  • For the write interconnect block, you can view the interconnect style, number of writes to the global memory, status of the -Xsforce-single-store-ring compiler flag, and number of store rings.
  • For the read interconnect block, you can view the interconnect style and number of reads from the global memory.
  • For the read interconnect router block, you can view the status of the -Xsnum-reorder flag, total number of buses, and all connections between buses and load LSUs. Buses in this block provide read data from the memory to load LSUs.
  • For the global memory (DDR in Figure 2), you can view the status of interleaving, interleaving size, number of channels, maximum bandwidth the BSP can deliver, and channel width.
  • For the memory controller block, you can view the maximum bandwidth the BSP can deliver, sum of the load/store throughput, and read/write bandwidth. For additional information about how global memory bandwidth use is calculated, refer to Global Memory Bandwidth Use Calculation in this guide. It describes the formulas used in calculating the bandwidth.
  • LSUs using USM pointers show up twice in both host and device global memory views as they can access both memories.

Block View

The block view of the System Viewer provides a more granular system view of the kernel. This view shows the following:

  • Fine-grained details within kernels (including instructions, dependencies, and schedule of the instructions) of the generated datapath of computations. The Intel® oneAPI DPC++/C++ Compiler encapsulates maximum instructions in clusters for better quality of results (QoR). The System Viewer shows clusters, instructions outside clusters, and their connections.
  • Linking from the instruction back to the source line by clicking the instruction node.
  • Various information about the instructions, such as data width, node's schedule information in the start cycle and latency, are provided if applicable.
NOTE:

The schedule information is relative to the start of each block. Because the Intel® oneAPI DPC++/C++ Compiler cannot statically infer the trip counts of blocks, if your design consists of multiple blocks, the compiler cannot compute the absolute schedule information by considering the trip counts. Moreover, the schedule information provides estimated values from empirical measurements for the stallable instructions (such as pipe RD/WR or memory LD/ST). The real schedule is likely to be different, and you must verify it with a hardware, or a simulation run.

If your design has loops, the Intel® oneAPI DPC++/C++ Compiler encapsulates the loop control logic into loop orchestration nodes and the initial condition of the loops into loop input nodes and their connection to the datapath.

Inside a block, there are often pipe RD/WR or memory LD/ST nodes connecting to computation nodes or clusters. You can click on the computation nodes and view the Details pane (or hover over the nodes) to see specific instructions and the bit width. You can click on the RD/WR or LD/ST nodes to see information such as instruction type, width, depth, LSU type, stall-free global memory, scheduled start cycle, estimated latency, and schedule of a pipe or an LSU from the Details pane. For stallable nodes, the latency value provided is an estimate. Perform a simulation or hardware run for more accurate latency values.

If your design has clusters, a cluster has a FIFO in its exit node to store any pipelined data in-flight. You can click on the cluster exit node to find the exit FIFO width and depth attribute. The cluster exit FIFO size is also available in the cluster view of the System Viewer.

Block View of the System Viewer Report

Cluster View

The cluster view of the System Viewer provides more granular graph views of the system or kernel. It helps in viewing clusters inside a block, and it shows all variables inside a cluster that have a loop-carried dependency. This view shows the following:

  • Fine-grained details within clusters (including instructions and dependencies of the instructions) of the generated datapath of computations.
  • Linking from the instruction back to the source line by clicking the instruction node.
  • Various information about the instructions, such as data width, node’s schedule information in start cycle, and latency, are provided if applicable.

A cluster starts with an entry node and ends with an exit node. The cluster exit node has a FIFO of depth greater than or equal to the latency of the cluster to store any data in-flight. You can find the size of the cluster exit FIFO by clicking on the exit node. The cluster exit FIFO size information is also available in the block view of the System Viewer when you click on the exit node.

Cluster View of the System Viewer

A cluster has a FIFO in its exit node to store any pipelined data in-flight. You can click on the cluster exit node to find the exit FIFO width and depth attribute. The cluster exit FIFO size is also available in the cluster view of the System Viewer.

Besides computation nodes, when your design contains loops, you can see loop orchestration nodes and variable nodes along with their Feedback nodes. The compiler generates the loop orchestration logic for loops in your design. Loop orchestration nodes represents this logic in the cluster view of the System Viewer. A variable node corresponds to a variable that has a loop-carried dependency in your design. A variable node goes through various computation logic and finally feeds to a Feedback node that connects back to the variable node. This back edge means that the variable is passed to the next iteration after the new value is evaluated. Scan for loop-carried variables that have a long latency to the Feedback nodes as they can be the II bottlenecks. You can cross-check by referring to the Loop Analysis report for more information about the II bottleneck. The Feedback node has a FIFO to store any data in-flight for the loop and is sized to d*II where d is the dependency distance, and II is the initiation interval. You can find the cluster exit FIFO size by clicking on the feedback node and looking at the Details pane or the pop-up box.

NOTE:

The dependency distance is the number of iterations between successive load/store nodes that depend on each other.

Parent topic: Review the FPGA Optimization Report
Level Two Title