Introduction
The Intel® System Bring-up Toolkit contains tools that enable developers to accelerate system bring-up and optimization—whether you’re a device manufacturer, system integrator, or IoT software provider. This page provides the release notes and documents the issues associated with the toolkit.
Intel® System Bring-up Toolkit includes the following components:
- Intel® VTune™ Profiler - Release Notes
- Intel® SoC Watch - Release Notes
- Intel® System Debugger - Release Notes
Version History
| Date | Version | Major Change Summary |
| September 2020 | 2021.1-beta09 | Intel® VTune™ Profiler: Performance Snapshot is now used as the default analysis type. Input and Output analysis is improved to capture PCIe MMIO reads and writes. GPU Accelerators: Added support for oneAPI Level Zero Specification for DPC++ Applications |
| July 2020 | 2021.1-beta08 |
Intel® System Debugger: Added capability to display Physical Memory in the Memory Viewer and Memory Browser. Added busy indicator for time consuming operations. Improved performance of Descriptor Table views and in-place register editing support Intel® VTune™ Profiler: Added Anomaly Detection analysis type in the Algorithm group. Added OpenMP Offload in HPC Analysis. Added Windows* Support for oneAPI Level Zero Specification for DPC++ Applications. Added demangling of DPC++ lambda function names. |
| June 2020 | 2021.1-beta07 |
Intel® System Debugger: Added samples for System Debug, System Trace and Intel(R) Debug Extensions for WinDbg*. Intel® VTune™ Profiler: GPU Accelerators - Support for oneAPI Level Zero Specification for DPC++ Applications; Update to IP Architecture diagram. New metrics in Input and Output Analysis. New TCP/IP Communication Agent. Cloud and Containerization - Display Container Name instead of Container ID |
| April 2020 | 2021.1-beta06 |
Intel® System Debugger: Transition to Python 3. Processor Trace support. Intel(R) Debug Extensions for WinDbg* improvements. Intel® VTune™ Profiler: System Overview - Hardware tracing improvements: represent module entry points, user/kernel metrics, interrupts. Bug fixes. |
| March 2020 | 2021.1-beta05 | Intel® VTune™ Profiler: Improvements and new analysis types for GPU accelerators. Platform analysis improvements. Intel® SoC Watch: Collection of tool usage analytics. New features and new platforms support. Error messages and help output improvements. Enhanced driver security. |
| February 2020 | 2021.1-beta04 | Support to profile DPC++ applications across CPUs and accelerators |
| November 2019 | 2021.1-beta03 | Initial Beta release of Intel® System Bring-up Toolkit |
New in This Release
- Intel® VTune™ Profiler
- Performance:
- Performance Snapshot Analysis Type as Default Analysis
- This release establishes Performance Snapshot as a starting point when you configure a new analysis. Previously, you started with the Hotspots analysis when you clicked Configure Analysis on the welcome screen. Performance Snapshot provides recommendations for next steps to help you select other analyses for deeper profiling.
-
Input and Output Analysis Improvements
-
The Input and Output analysis type features a new methodology for locating sources of reads and writes targeting Memory-Mapped I/O (MMIO) address space regions to which I/O devices are mapped. Such MMIO reads and writes are expensive loads and stores resulting in Outbound PCIe traffic.
-
- Performance Snapshot Analysis Type as Default Analysis
- GPU Accelerators:
- Support for oneAPI Level Zero Specification for DPC++ Applications
- Intel® VTune™ Profiler now supports version 1.0.4 of the oneAPI Level Zero API specification when you run GPU analyses (GPU Offload analysis and GPU Compute/Media Hotspots) on DPC++ applications in Windows* and Linux* environments.
- Support for oneAPI Level Zero Specification for DPC++ Applications
- Support for DPC++ Applications:
- Refer to Intel® VTune™ Profiler Release Notes for the complete list of new features
- Performance:
- Intel® System Debugger
- No new features
- Intel® SoC Watch
- No new features
- Intel® VTune™ Profiler
- Algorithm group:
- This release introduces the Anomaly Detection analysis type in the Algorithm group. Use this analysis to detect performance anomalies in frequently recurring code intervals including loop iterations. Anomaly Detection uses Intel® Processor Trace (Intel® PT) technology to perform detailed analysis at the microsecond level.
- Parallelism:
- The HPC Performance Characterization analysis type now supports the offload of OpenMP regions. The support includes these additions to the analysis type:
- The summary pane now includes a breakdown of OpenMP offload time by Compute, Data Transfer, and Overhead. You can also see a table with the top 5 OpenMP target regions sorted by offload time.
- The bottom-up pane now allows grouping by OpenMP Offload Region. With this grouping active, the grid displays several new columns. The timeline shows scale markers that indicate the span of OpenMP offload regions and OpenMP operations internal to those regions.
- The HPC Performance Characterization analysis type now supports the offload of OpenMP regions. The support includes these additions to the analysis type:
- GPU Accelerators:
- Windows* Support for oneAPI Level Zero Specification for DPC++ Applications: This release extends support for the oneAPI Level Zero API specification to Windows* systems when you run GPU analyses (GPU Offload analysis and GPU Compute/Media Hotspots) on DPC++ applications on these systems. Previously, support for this specification existed on Linux* systems only. For these DPC++ applications, Intel® VTune™ supports version 0.91.10 of the oneAPI Level Zero API.
- Issue Markers in Memory Hierarchy Diagram: The Memory Hierarchy Diagram of the GPU Compute/Media Hotspots analysis now displays the same markers to highlight metrics as the ones used to indicate performance or data issues in the Summary and Grid displays. This provides a consistent look and feel to the diagram and helps you correlate metrics between both displays.
- Support for DPC++ Applications:
- Demangling of Lambda Functions: This release implements the demangling of DPC++ lambda function names, which are used as DPC++ kernel names.
- Refer to Intel® VTune™ Profiler Release Notes for the complete list of new features
- Algorithm group:
- Intel® System Debugger
- Added capability to display Physical Memory in the Memory Viewer and Memory Browser.
- Added a busy indicator on the status bar which will be visible for time consuming operations.
- Added an option in Eclipse to perform a Power-Good reset of the target, to mimic a power-cycle.
- Added a feature to quickly setup a breakpoint at the entry of modules.
- Added support for the protection keys for supervisor-mode pages (CR4.PKS and the IA32_PKRS MSR).
- Improved performance of Descriptor Table views and in-place register editing support.
- Intel® SoC Watch
- No new features
- Intel® VTune™ Profiler
- GPU Accelerators:
- Support for oneAPI Level Zero Specification for DPC++ Applications
- Update to IP Architecture diagram: The IP Architecture Diagram of the GPU Compute/Media Hotspots analysis has been renamed to Memory Hierarchy Diagram. The diagram has been re-designed to improve the look and feel, and can help make the understanding of metrics more intuitive.
- New TCP/IP Communication Agent:
- This release features a new TCP/IP communication agent as a connection type, intended for profiling embedded systems running real-time operating systems
- Cloud and Containerization
- This release displays Container Name instead of Container ID to simply container’s identification
- Refer to Intel® VTune™ Profiler Release Notes for the complete list of new features
- GPU Accelerators:
- Intel® System Debugger
- Added Samples, providing use cases examples for System Debug, System Trace and Intel(R) Debug Extensions for WinDbg*.
- UX Improvement: proper display of platform break hit events in the Debug view
- Added platform breaks related to virtualization
- Added logical thread filtering support based on regular expressions
- Increased the default search range of "Load available debug symbols" feature.
- Added support for TSXLDTRK instruction.
- Added support for SERIALIZE instruction for lighter-weight serialization.
- Added an error message indicating when the Intel® Processor Trace (Intel® PT) or Last Branch Record (LBR) buffers are empty.
- Intel® SoC Watch
- No new features
- Intel® VTune™ Profiler
- System Overview :
- Hardware tracing Improvements: represent module entry points, user/kernel metrics, interrupts
- User Interface improvements:
- Proper handling of too short collections for PMU-based analysis
- Refer to Intel® VTune™ Profiler Release Notes for the complete list of new features
- System Overview :
- Intel® System Debugger
- Eclipse* IDE updated to 2020-03
- Transitioned to Python* 3
- Added Target Indicator and Telemetry
- System Debug: Added Processor Trace support
- System Trace: Added ENUM feature
- Intel® Debugger Extension for WinDbg*: Added support for ACPI Machine Language debugging in
- Intel® Debugger Extension for WinDbg*: Added support for Windows* Core OS target in
- Intel® SoC Watch
- No new features
- Intel® VTune™ Profiler
- GPU accelerators support:
- GPU Compute/Media Hotspots analysis in the Dynamic Instruction Count mode has been extended to include SIMD utilization metrics at the kernel and instruction level.
- Deeper GPU utilization analysis has been introduced in Application Performance Snapshot (APS) and the HPC Performance Characterization analysis.
- Application Performance Snapshot GPU compute metrics include OpenMP* offload efficiency metrics.
- Simplified dependency on the Intel® Metrics Discovery Application Programming Interface library to collect GPU hardware statistics on Linux* systems.
- Platform analysis improvements:
- CPU/FPGA Interaction analysis has been extended to process data sources collected either with AOCL Profiler (new mode) and via OpenCL™ Profiling API (legacy mode).
- The Hardware Tracing mode in the System Overview analysis has been extended to include new metrics to make the analysis more kernel-aware.
- Refer to Intel® VTune™ Profiler Release Notes for the complete list of new features
- GPU accelerators support:
- Intel® System Debugger
- No new features
- Intel® SoC Watch
- Added collection of tool usage analytics
- Added new features pch-slps0, pch-slps0-dbg
- Improved error messages and help output. Enhanced driver security
- Added support for Intel platform code named Comet Lake
- Added support for Intel platform code named Ice Lake
- Intel® VTune™ Profiler: Profile DPC++ code across CPUs and multiple accelerator architectures, including GPUs and FPGAs on Linux and Windows OS. Refer to Intel® VTune™ Profiler Release Notes for the complete list of new features
- Intel® System Debugger: No new features
- Intel® SoC Watch: No new features
- Intel® System Debugger started the transition phase from Python* 2.7 to Python 3. The transition to Python 3 will be finalized by end of 2019. The debugger ships both Python 2.7 and Python 3.6 versions during the transition phase until 2020.
- New platform support for Intel® System Debugger:
- 10th generation Intel® Core™ processor (Ice Lake / Ice Lake PCH-LP)
- 10th generation Intel® Core processor (Comet Lake U / Comet Lake PCH-LP)
- 10th generation Intel® Core processor (Amber Lake Y 4+2 / Sunrise Point PCH-LP)
- Intel® Xeon® processor (Cascade Lake / Lewisburg PCH)
System Requirements
For software and hardware requirements, please refer to Intel® System Bring-up Toolkit System Requirements.
Installation Instructions
The Intel® System Bring-up Toolkit can be downloaded and installed from the download page. Please visit Installation Guide for Intel® oneAPI Toolkits for more detailed instructions.
Linux Installation
Step 1: Use the following command to launch the GUI installer as root:
sudo ./l_BringupKit_[version].sh
Step 2: Follow the installer instruction screens.
- Check that /etc/udev/rules.d/99-dci.rules file exists after the installation.
Note: Root permissions or sudo is required to create this file.
Windows Installation
Step 1: Double-click the EXE file to launch the GUI installer.
w_BringupKit_[version].exe
Step 2: Follow the installer instruction screens. Once the installation is complete, verify that your toolkit has been installed to the correct installation directory, as follows:
| Linux | Windows |
|---|---|
/opt/intel/oneapi |
C:Program Files (x86)InteloneAPI |
Getting Started
Known Issues and Workarounds
- When using Samples in Eclipse* IDE: The “Problems” view in Eclipse* may list error results for some oneAPI sample projects. These errors are false positives and can be ignored. The samples will build and run in Eclipse when the environment is correctly configured as described in the oneAPI Getting Started documents.
For component specific issues please check their respective Release notes
Notices and Disclaimers
Intel technologies may require enabled hardware, software or service activation.
No product or component can be absolutely secure.
Your costs and results may vary.
© Intel Corporation. Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries. Other names and brands may be claimed as the property of others.
No license (express or implied, by estoppel or otherwise) to any intellectual property rights is granted by this document.
The products described may contain design defects or errors known as errata which may cause the product to deviate from published specifications. Current characterized errata are available on request.
Intel disclaims all express and implied warranties, including without limitation, the implied warranties of merchantability, fitness for a particular purpose, and non-infringement, as well as any warranty arising from course of performance, course of dealing, or usage in trade.