Intel® System Studio Release Notes, System Requirements, and What's New

 

This page provides system requirements and release notes for Intel® System Studio. They are are categorized by year, from the newest to oldest, with individual releases listed within each year. For component-specific system requirements and release notes, please see the Release notes for individual components section below.

Click a version to expand it into a summary of new features and changes in that version since the last release, and access the download buttons for the detailed system requirements and release notes, which include important information, such as pre-requisites, software compatibility, installation instructions, and known issues.

2020

Update 3

Release Notes  System Requirements

• Intel® System Studio 2020 Update 3 includes functional and security updates. Users should update to the latest version.
• Intel® Integrated Performance Primitives (Intel® IPP):
  • Extends optimization for Intel® IPP CRC functions on 10th Generation Intel® Core™ processor family.
  • Extends optimizations for Intel® IPP Image Processing and Color Conversions functions for selected algorithms for Intel® Advanced Vector Extensions 512 (Intel® AVX-512).
  • Extends Intel® IPP Cryptography with ECDSA (Sign) and ECDHE for the NIST curves p256r1 and p384r1.
• Intel® VTune™ Profiler:
  • Adds support for 11th Generation Intel® Core™ processors codenamed Tiger Lake, including hotspots, microarchitecture exploration, memory access, and GPU analyses.
  • I/O analysis: source level Memory Mapped I/O (MMIO) supports InfiniBand devices.
  • Profiling of applications annotated with Intel® Instrumentation and Tracing Technology (ITT) APIs is enhanced with Average Task Time and Average Frame Time metrics.
  • Adds support for remote profiling of applications running in Amazon Web Services* (AWS) EC2 instances.

 

Update 2

Release Notes System Requirements

• Intel® System Studio 2020 Update 2 includes functional and security updates. Users should update to the latest version.
• Intel® Integrated Performance Primitives (Intel® IPP):
  • Adds new universal CRC function to compute CRC8, CRC16, CRC24, CRC32 checksums.
  • Enables the Intel® IPP Threading Layer for the ippiCrossCorr function.
  • Reinstates the ippiComplement function optimized for Intel® Advanced Vector Extensions 512 (Intel® AVX-512), Intel® Advanced Vector Extensions 2 (Intel® AVX2), Intel® Streaming SIMD Extensions 4.2 (Intel® SSE4.2) instruction sets.
  • Extends ippsHilbert_32f32fc on double precision counterparts.
  • Adds IPP Custom Library Tool custom dispatcher support.
• Intel® Math Kernel Library (Intel® MKL):
  • BLAS: Introduces {cblas_}?axpy_batch APIs. Introduced bfloat16 data type support for GEMM and pack-API. Fixes a parameter validation error of cblas_zgemmt to allow CblasConjTrans value.
  • LAPACK: Improves performance of {D,S}GESDD for sequential mode.
  • ScaLAPACK: Aligns Intel MKL ScaLAPACK functionality with Netlib ScaLAPACK 2.1.0. Adds robust ScaLAPACK routines for computing the QR factorization with column pivoting.
  • Vector Statistics: Improves performance for the threading version of Sobol quasi-random number generator in case of registered user-defined parameters.
  • Graph: Significant API changes to enable better consistency and uniformity. Adds new API for transposing graph data. Adds support for sparse vectors and Compressed Sparse Column (CSC) matrix format. Adds support for vxm. Adds support for new semirings and extended support for descriptor flags which can be used for Connected Components(CC), Triangle Count (TC), Betweenness Centrality (BC) and Breadth First Search (BFS). Adds PLUS accumulator for mxv which can be used for PageRank (PR).
  • Vector Math: Introduces strided API.
• Intel® Threading Building Blocks (Intel® TBB):
  • Changes body type concept of the flow::input_node. Set TBB_DEPRECATED_INPUT_NODE_BODY to 1 to compile with the previous concept of the body type.
  • Fixes compilation errors in C++20 mode due to ambiguity of comparison operators.
  • Fixes an issue in TBBBuild.cmake that causes the build with no arguments to fail.
  • Adds cmake/{TBBConfig,TBBConfigVersion}.cmake to Git ignore list.
• Intel® Data Analytics Acceleration Library (Intel® DAAL):
  • Introduces Thunder method for Support Vector Machine (SVM) training algorithm, which demonstrates better training time than the existing sequential minimal optimization method.
  • Extends training with the number of features greater than the number of observations for Linear Regression, Ridge Regression, and Principal Component Analysis.
  • Adds new sample_weights parameter for SVM algorithm.
  • Adds new parameter in K-Means algorithm, resultsToEvaluate, which controls computation of centroids, assignments, and exact objective function.
  • Improves performance of Support Vector Machine training and prediction, Elastic Net and LASSO training, Principal Component Analysis training and transform, K-D tree based k-Nearest Neighbors prediction.
  • Improves performance of K-Means algorithm in batch computation mode.
  • Improves performance of RBF kernel function.
  • daal4py library: Improves performance of pandas input format. Improves Scikit-learn compatible API for AdaBoost classifier, Decision Tree classifier, and Gradient Boosted Trees classifier and regressor.
  • Improves performance of the following Intel Scikit-learn algorithms and functions: fit and prediction in K-Means and Support Vector Classification (SVC), fit in Elastic Net and LASSO, fit and transform in PCA; Support Vector Classification (SVC) with non-default weights of samples and classes; train_test_split() and assert_all_finite().
• Intel® Advisor adds Memory-Level Roofline feature (previously known as Integrated Roofline, tech preview feature).
• Intel® VTune™ Profiler:
  • Adds Performance Snapshot analysis type to get a quick overview of issues that affect your application performance. Performance Snapshot characterizes the workload on the system and provides recommendations for next steps to help select other analyses for deeper profiling.
  • Platform I/O metrics can now be attributed to individual devices managed by Intel® VMD technology.
  • Enhances I/O Analysis for Sky Lake and Cascade Lake servers by highlighting code that potentially performs MMIO reads.
  • Adds support for Intel processors code named Cooper Lake and Comet Lake.
  • Adds support for Ubuntu* 20.04, Fedora* 32, and Microsoft* Windows* 10, May 2020 Update.
• Intel® System Debugger:
  • Adds Intel® Processor Trace (Intel® PT) support - DRAM-based execution tracing with a larger execution history compared to the method based on Last Branch Record (LBR).
  • Enables loading of debug symbols for stack frames and thread nodes and changes the "Load available debug symbols" (formerly LoadThis) button enabling conditions accordingly.
  • Enables loading of the UEFI DXE modules and adds a dedicated "Load EFI Images" button (also known as LoadDXE in System Debug Legacy).
  • Adds dedicated button to load all Pre-EFI Initialization (PEI) phase modules (also known as LoadPEIMs in System Debug Legacy).
  • Introduces a single menu for all breakpoint types: address, IO port access, generic, C/C++ breakpoints, and platform breaks. The menu can be accessed from the Breakpoints view using the "Add Breakpoint" button or selecting "Add System Debug Breakpoint" from the View Menu.
  • Adds support for additional platform breakpoints including VM-Launch, VM-Clear, and VM-Exit.
  • Improves Debug View to display platform break hit events.
  • Adds support to display IA descriptor tables (GDT, LDT and IDT).
  • Adds logical thread filtering support based on regular expressions.
  • System Debug User Guide documents instructions for loading symbols for CoreBoot*.
  • Intel® Debug Extensions for WinDbg* adds support for Microsoft Windows 10* (20H1) May 2020 update.
  • The Android* Trace feature of Intel® System Debugger - System Trace is permanently removed from Intel® System Studio, beginning with this release.

Update 1

Release Notes System Requirements

• Intel® System Studio 2020 Update 1 includes functional and security updates. Users should update to the latest version.
• Updates base Eclipse* IDE to Eclipse IDE 2019-12 and includes functional and security updates.
• Improves Yocto Project* documentation.
• Intel® Math Kernel Library (Intel® MKL):
  • Improves BLAS GEMM3M performance on Intel® Advanced Vector Extensions 2 (Intel® AVX2) and Intel® Advanced Vector Extensions 512 (Intel® AVX-512) architectures.
  • Improves performance of LAPACK {D,S}GESDD for case jobz='N'.
  • Introduces ScaLAPACK distributed nonsymmetric eigensolver functionality (P?GEEVX) for computing eigenvalues and optionally eigenvectors for a general nonsymmetric matrix.
  • Improves performance for FFTs with MKL TBB threading layer on CPUs.
  • Introduces Graph functionality as a preview feature that supports the sparse linear algebra operations and semirings used in PageRank and Triangle Count algorithms.
• Intel® Integrated Performance Primitives (Intel® IPP):
  • Extends Intel® IPP Image Processing ippsFIRSparse*_32fc functions optimizations for Intel® Advanced Vector Extensions 2 (Intel® AVX2) and Intel® Advanced Vector Extensions 512 (Intel® AVX-512).
  • Optimizes Intel® IPP Image Processing Resize 8u functions for 10th Generation Intel® Core™ Processor and ippsAddProduct_32fc for Intel® Advanced Vector Extensions 2 (Intel® AVX2) and Intel® Advanced Vector Extensions 512 (Intel® AVX-512).
  • Introduces Intel® IPP Cryptography RSA multi-buffer cipher API to process up to 8 messages simultaneously. The API is optimized for 10th Generation Intel® Core™ processor family with performance improvement up to 6x for decryption and up to 4x for encryption compared to the RSA classic single-buffer API.
• Intel® Advisor:
  • Advisor Python* API uses Python 3.
  • Adds First Bottleneck visualization for Integrated Roofline chart; and integrates Roofline single kernel view and guidance.
• Intel® SoC Watch:
  • Adds (user opt-in) collection of SoC Watch usage analytics.
  • Adds reporting residency in lowest power sleep state, SLP_S0 (s0i3/Modern Standby).
  • Adds reporting blocking reasons (IP state conditions) that may be preventing entry to lowest power sleep state, SLP_S0 (s0i3/Modern Standby).
  • Enhances SoC Watch driver security by limiting access to only registers or memory required to collect platform power and performance information.
• Intel® VTune™ Profiler:
  • Adds Microarchitecture Exploration analysis support on Intel processors codenamed Icelake.
  • Extends GPU Compute/Media Hotspots analysis in the Dynamic Instruction Count mode to include SIMD utilization metrics at the kernel and instruction level, which help identify instructions in the OpenCL™ kernel that utilize SIMD poorly.
  • Adds a deeper GPU utilization analysis in Application Performance Snapshot (APS) and the HPC Performance Characterization analysis. The GPU utilization analysis now includes GPU time, GPU IPC, GPU utilization, and % of stalled and idle execution units.
  • VTune Profiler automatically selects the latest libstdc++ available in runtime to satisfy the GPU analysis requirements on Linux* systems.
  • Extends CPU/FPGA Interaction analysis to process data sources collected either with AOCL Profiler (new mode) or via OpenCL Profiling API (legacy mode). Specify the name of your application target and its parameters directly in the WHAT pane.
  • Extends the Hardware Tracing mode in the System Overview analysis to include new metrics to make the analysis more kernel-aware. The CPU Time metric has been split to User Time and Kernel Time metrics to show the number of kernel mode switches and their frequency (switches per second). OS Kernel Activity and OS Scheduling metrics identify anomaly issues caused by unexpected kernel activity or preemptions
  • Adds a Module Entry Point grouping level to the System Overview viewpoint to display data collected in the Hardware Tracing mode. The grouping shifts the focus to precise CPU time spent within system calls, interrupts, or within a runtime library API.

Initial Release

Release Notes System Requirements

• Extends Intel® C++ Compiler support of C++ 17 and provides initial C++ 20 language standard support.
• Intel® SoC Watch adds support for Intel platform code named Comet Lake and Intel platform code named Ice Lake.
• Renames Intel® VTune™ Amplifier to Intel® VTune™ Profiler starting with version 2020.
• Intel® VTune™ Profiler:
  • Adds new GPU Offload analysis to explore and correlate code execution across CPUs and GPUs, and identify a kernel of interest for GPU-bound applications to be explored further with GPU Compute/Media Hotspots analysis.
  • Updates GPU Compute/Media Hotspots analysis with options for GPU in-kernel analysis.
  • Extends command line hotspots report scope to support GPU analysis types.
  • Improves dynamic instruction count collection available as part of the GPU Compute/Media Hotspots analysis to provide better accuracy for basic block assembly analysis.
  • Updates System Overview analysis to serve as an entry point to platform analysis assessing system (IO, accelerators and CPU) performance and providing guidance for further analysis steps, and adds New Hardware Tracing mode.
  • Adds new Throttling analysis to identify causes for system throttling, like exceeding safe thermal or power limits.
  • Extends Energy analysis options to monitor processor package energy consumption over time and identify how it correlates with CPU throttling.
  • Extends containerization support with an option to install and run the VTune Profiler in a Docker* container and profile targets both inside the same container as well as outside the container.
• Adds Max and Bound Bandwidth metrics to Application Performance Snapshot to better estimate the efficiency of the DRAM, MCDRAM, Persistent Memory and Cornelis Networks usage.
• Platform Profiler:
  • Extends Overview and Memory views with new metrics for analyzing Non Uniform Memory Access (NUMA) behavior.
  • Implements user authentication and authorization to enable access control to users’ data.
  • Adds a new option for users to choose or modify the location of Platform Profiler data files.
• Updates Intel® System Debugger to a new Eclipse*-integrated source-level debugger, providing reworked support for platform registers, improved support for PCI configuration space, better debug information, and a Python*-based scripting console.
• Adds Intel® System Debugger Target Indicator, a cross-platform tool that indicates the status of a DCI debug connection to a target platform.
• Adds Intel® System Debugger support for:
  • 10th Generation Intel® Core™ Processor (Ice Lake) / Ice Lake PCH-LP
  • 10th Generation Intel® Core™ Processor (Comet Lake) / 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
• OpenCL™ API Debugger and Kernel Development Framework functionality is deprecated in this release. Intel® VTune™ Profiler can be configured to see OpenCL™ API calls. Intercept Layer for OpenCL™ Applications can be used for replacement of OpenCL™ API Debugger functionality.
• Adds OpenCL™ tools support for:
  • 10th Generation Intel® Core™ Processor named Ice Lake
  • Microsoft Visual Studio* 2019
  • CentOS* 8
  • Adds OpenCL™ tools offline compiler support for generation of optimized ELF binary file from SPIR-V file
  • Bug fixes and security updates
• To support building OpenVINO™ applications, Intel® System Studio 2019 provides instructions for the user to create a custom Docker container that contains OpenVINO tools and libraries. Intel® System Studio 2020 includes a Docker platform to build OpenVINO projects as a user-selectable option. This Docker platform contains OpenVINO 2019 R3.1 release.
• Deprecates Ubuntu* 16.04 Docker files for remote IoT targets. Users are recommended to start with Ubuntu 18.04.
• Removes MRAA/UPM binaries and target update capability from IoT Connection Tools since they are now distributed by the Eclipse* Foundation. However, the Sensor Explorer feature remains functional.
• Updates IoT and Cloud code samples. Removes IoT code samples that use sensors from discontinued developer kits.
• Supports latest IoT developer kits, including accelerator card release updates to the IEI AIoT Tank Developer Kit.
• Updates base Eclipse* IDE to Eclipse IDE 2019-09.
• Updates JDK* to 11.0.5 which includes CVE fixes.

2019

Update 5

Release Notes FreeBSD* Target Release Notes System Requirements

• Intel® System Studio 2019 Update 5 includes functional and security updates. Users should update to the latest version.
• Intel® C++ Compiler includes a new -qnextgen compiler option that uses LLVM Technology. For more details, see the additional requirements for ICC NextGen and the ICC NextGen notes sections in the Intel® C++ Compiler release notes for Linux* host, Windows* host, and Windows* target.
• Intel® VTune™ Amplifier 2019 Update 5 enhancements include:
  • Interactive help tour available from the welcome page guides users through the product interface using a sample project.
  • Enables automatic password-less access configuration to a remote Linux* (SSH) target on Windows*.
  • Relaxes limitations on collecting GPU hardware metrics for users without administrator / root privileges.
  • Relaxes limitations for microarchitecture exploration analysis with Perf* driverless collection on Linux for users with perf_event_paranoid set to 2 to support event-based sampling in user space.
  • Updates memory access analysis with a new UPI utilization metric for Intel microarchitectures code named Cascade Lake and Skylake.
  • Adds hardware-based analysis support for systems running on Hyper-V.
  • Adds Application Performance Snapshot max and bound metrics to estimate the efficiency of the DRAM, MCDRAM, and persistent memory usage.
• Intel® Threading Building Blocks 2019 Update 7 and 8 enhancements include:
  • Improved Intel(R) Threading Building Blocks (TBB) async_node to never block a thread that sends a message through its gateway.
  • Adds Windows* support to the CMake module TBBInstallConfig.
  • Adds ordered associative containers preview feature: concurrent_{map,multimap,set,multiset} (requires C++11).
  • Fixes enqueuing tbb::task into tbb::task_arena not to fail on threads with no task scheduler initialized.

Update 4

Release Notes FreeBSD* Target Release Notes System Requirements

• Intel® System Studio 2019 Update 4 does not include the latest functional and security updates. Intel System Studio 2019 Update 5 is targeted to be released in September 2019 and will include additional functional and security updates. Users should update to the latest version as it becomes available.
• Intel® C++ Compiler adds Microsoft Visual Studio 2019* support.
• Enhances Intel® SoC Watch -f hw-cpu-pstate feature reporting to include Core P-state/Frequency Histogram summary, which makes it easier to see residencies in turbo and throttled frequency states.
• Adds Intel® System Debugger support for:
  • Intel® Xeon® E Processor (Coffee Lake-S) Server
  • Intel® Pentium® Gold processors: 5405U (WHL-U platform based)
  • 4417U (KBL-R platform based), Intel® Celeron® processors: 4205U (WHL-U platform based), 3867U (KBL-R platform based)
  • 9th Gen Intel® Core™ Processor: i9-9900KF, i7-9700KF, i5-9600KF, i5-9400, i5-9400F, i3-9350KF
  • Intel® Core™ X-series Processor (Basin Falls Refresh): i9-9980XE, i9-9960X, i9-9940X, i9-9920X, i9-9900X, i9-9820X, i7-9800X.
• Adds "Control-flow Enforcement Technology (CET)" debug support for GDB server.
• Updates Intel System Studio's IDE for Linux* targets with Eclipse* 2019-03 support.

Update 3

Release Notes FreeBSD* Target Release Notes System Requirements

• Intel® System Studio 2019 Update 3 includes functional and security updates. Users should update to the latest version.
• Updates Intel System Studio's IDE for Linux* targets with Java 11 and Eclipse* 2018-12 support.
  • Adds Java* 11 JDK to simplify getting started with Java development.
  • Updates Java* samples for Java 11
• Adds Intel® VTune™ Amplifier support for Intel® Optane™ DC persistent memory and the latest microarchitecture code-named Cascade Lake. This includes new hardware event support and enhanced memory analysis to design and optimize for the new persistent memory technology.
• Enhances VTune Amplifier PCIe device metrics for I/O traffic in the Input and Output analysis to help understand interactions between Cores and Network Interface Cards (NICs).
• Enhances Intel® SoC Watch informational messages for CPU C-State reports. Improves organization of SoC Watch reports in the trace results file generated by "-r int".
• Removes limitation on the number of SoC Watch metrics that can be imported into VTune Amplifier for visualization.
• Improved Intel® Advisor Summary pane with a sleeker look and better program-level issue highlighting, including a new block with memory bandwidth information. Survey columns can be customized using the column configurator. Сache Simulator configuration dialog allows visual configuration of different cache levels for Memory Access Patterns analysis. Improved Roofline comparison visually distinguishes the compared result sets.
• Intel® C++ Compiler adds value safe simd options for #pragma omp simd, new custom memory allocator library, and supports new code names in -[Q]x / -[Q]ax / -[m]tune / -[m]arch options. See details.
• Updates GNU* GDB-IA CDT plug-in to support Java* 11 and Eclipse* 2018-12. Native GDB-IA supports "Shadow Stack" and "Indirect branch tracking" capabilities of "Control-flow Enforcement Technology (CET)."

Update 2

Release Notes FreeBSD* Target Release Notes System Requirements

• Intel® System Studio 2019 Update 2 includes functional and security updates. Users should update to the latest version.
• Enhances Intel® VTune™ Amplifier configuration for Microarchitecture Exploration analysis to provide control over collected hardware metrics and data collection overhead. Supports managed Linux* and Windows* targets with tiered compilation for .NET* Core 3.0 Preview 1 and .NET Core 2.2.
• Adds System Analyzer tool for monitoring real-time metrics on a target system to VTune Amplifier as a PREVIEW feature. See the VTune Amplifier Performance Analysis Cookbook recipe for more details.

Update 1

Release Notes FreeBSD* Target Release Notes System Requirements

• Adds Python* support for easily developing native Python-based edge application using the Eclipse* IDE and the integrated Pydev* plug-in.
• Extends Intel® VTune™ Amplifier threading analysis with the lower overhead hardware event-based sampling mode for analyzing the impact of thread pre-emption and context switching.
• Expands analysis for OpenCL™ applications using Intel® VTune™ Amplifier. OpenCL™ Tools now deliver a bundled CPU runtimes.
• Provides flexibility and ease of building computer vision projects within Docker containers locally or remotely. Get started seamlessly using samples from Intel® Distribution of OpenVINO™ toolkit.
• Extends the Library Explorer to more easily browse optimized library functions (math, data analytics, threading, signal / image, and more).
• Supports additional hardware development kits: IEI Tank AIoT Developer Kit (IEI Tank 870-Q170) and the UP Squared* AI Vision Developer Kit (RE-UPS-APLC2-A10-0232).

Initial Release

Release Notes FreeBSD* Target Release Notes System Requirements

Overview:

• Rapid insights into platform-level performance. New work flow and simplified set up make Intel® VTune™ Amplifier easier to use. Get fast insights into overall system configuration, performance, and behavior, with specific focus on identifying platform-level memory and storage bottlenecks and imbalances.
• Code vectorization and threading. New Intel® Advisor tool provides vectorization optimization and threading design tools to ensure applications take full performance advantage of hardware capabilities.
• OpenCL™ tools help offload compute to Intel® processors (CPU and GPU / Intel® Processor Graphics). Streamline building, debugging, and analyzing OpenCL™ applications through new projects, or customize kernel code from Intel computer vision and media software tools.
• Improve debug. Intel® Processor Trace, with timestamps over direct connect interface (DCI), helps triage timing and performance issues with Intel® Debug Extensions for WinDbg*. Easily set up the target and connection type, then manage multiple simultaneous debug or trace connections to quicken system development. Configure and automate system trace to run on the command line.
• Create a cloud project. Use sample projects for connecting to cloud services providers, including Microsoft Azure*, Amazon Web Services*, Google Cloud IoT Core* (new!), and IBM Bluemix.
• Build Yocto Project* compatible applications and platform projects, for the Yocto Project compatible targets Wind River Linux* LTS 17 (and above) or Yocto Project 2.6 (and above), using the Intel® System Studio Eclipse IDE and the Yocto Project Toaster tool on Linux* development hosts.
• New download and install pages guide you through selecting and installing the Intel® System Studio tools and libraries that you need.
• Improved ease of use with more than 65 new code wizards and samples.
• Harness more performance with the latest updates of the performance libraries and analysis and debugger tools with new and enhanced features, and continued support for the latest Intel® platforms.
• Hardware support: 8th Generation Intel® Core™ U- and S-Series Processors, and Intel® Xeon® Processor D-1500 product family.

Detailed New Features List

Intel® VTune™ Amplifier

• Simplified analysis configuration workflow provides pre-selected target and collection options available in the same view.
• Default finalization mode set to Fast to minimize post-processing overhead if the number of collected samples exceeds the threshold.
• New Hotspots analysis combines former Basic Hotspots and Advanced Hotspots analyses configurations and is extended to show additional performance insights such as hardware usage efficiency and vector register utilization.
• New Threading analysis combines and replaces former Concurrency and Locks and Waits analysis types.
• New Platform Profiler provides low-overhead, system-wide analysis and insights into overall system configuration, performance, and behavior. Use this tool to:
  • Identify bottlenecks by monitoring over- or under-utilized subsystems and buses (CPU, storage, memory, PCIe, and network interfaces) and platform-level imbalances.
  • Understand a system topology using diagrams annotated with performance data.
• Microarchitecture Exploration analysis improvements
  • Microarchitecture Exploration analysis (formerly General Exploration) configuration was split to provide options for a lightweight summary or full details analysis with all levels of PMU metrics.
  • Microarchitecture Exploration analysis view was extended with the µPipe hardware metric representation to easily identify bottlenecks in hardware usage and provide quick insights for optimization.
• Input and Output (I/O) analysis on Linux* extended to profile DPDK and SPDK I/O API. Use this data to correlate CPU activity with network data plane utilization, visualize PCIe bandwidth utilization per NIC, estimate UPI bandwidth on multi-socket systems, and identify bottlenecks.
• Containerization flexibility and support improvements
  • Adds user-mode sampling and tracing collection (Hotspots analysis) support for Docker* container targets.
  • Adds profiling support for targets running in the Singularity* containers.
  • Adds profiling for native and Java* applications in the Docker and LXC containers.
• Managed runtime analysis improvements
  • Extends JIT profiling for server-side applications running on the LLVM* or HHVM* PHP servers to support the event-based sampling analysis in the attach mode.
  • Extends Java code analysis with support for OpenJDK* 9 and Oracle* JDK 9.
  • Enables Hotspots analysis in the hardware event-based sampling mode for .NET* Core applications on Linux and Windows* systems for the Launch Application target type.
• Embedded Platforms and Accelerators analysis improvements
  • New CPU/FPGA Interaction analysis (PREVIEW) to assess the balance between the CPU and FPGA on systems with a discrete Intel® Arria® 10 FPGA running OpenCL™ applications.
  • New Graphics Rendering analysis (PREVIEW) for CPU/GPU utilization of code running on the Xen* virtualization platform installed on a remote embedded target.
  • Supports sampling command-line analysis on remote QNX* embedded systems via Ethernet connection.
• KVM guest OS profiling was extended to profile both KVM kernel and user space from the host system for a full-scale performance analysis of host and VM systems.
• Updated GPU Hotspots analysis on Linux to use Intel Metric Discovery API library for GPU metric collection, which involves support for kernel 4.14 and higher.
• Application Performance Snapshot improvements
  • Adds uncore-based metrics for DRAM/MCDRAM memory analysis to help identify whether an application is bandwidth bound.
  • Adds the ability to pause/resume collection with MPI_Pcontrol and ITT API. Adds -start-paused option to exclude application execution from collection from the start to the first collection resume occurrence.
  • Enables selection of which data types are collected to reduce overhead (MPI tracing, OpenMP tracing, hardware counter based collection).
  • Exposes the CPU Utilization metric by physical cores on processors that support proper hardware events.
  • Significantly reduces MPI tracing overhead when there are a large number of ranks.
  • Improves integration with Intel® Trace Analyzer and Collector by adding the ability to generate profiling configuration files with the aps-report option.
• Hardware event-based analysis supported for targets running in the Hyper-V* environment on Windows* 10 Fall Creators Update (RedStone3).
• Adds HTML and XML to command line output formats.
• Supports newer versions of several OSs: SUSE Linux Enterprise Server* (SLES) 12 SP3 and 15, Red Hat Enterprise Linux* 7.5, Fedora* 28, Ubuntu* 18.04, Microsoft Windows* 10 RS4

Energy Analysis / Intel® SoC Watch

• Supports Intel platforms codenamed Whiskey Lake and Amber Lake.
• Adds option to delay start of a workload (program-delay).
• Adds new group names to make it easier to find and select features.
• The -f hw-cpu-pstate summary report includes a new core P-state average frequency table containing the core P-state approximate average and actual average frequency.

       Intel SoC Watch for Linux* / Android*

       Intel® SoC Watch for Windows*

• Adds guard against sampling bandwidth data too frequently on certain platforms.
• Adds a new energy metric.

Intel® Advisor -- New Tool!

• Vectorization Advisor is a vectorization analysis tool that identifies loops that may benefit most from vectorization, and what is blocking effective vectorization; explores the benefit of alternative data reorganizations; and increases the confidence for safe vectorization.
• Threading Advisor is a threading design and prototyping tool that lets you analyze, design, tune, and check threading design options without disrupting your normal development.
• Optimize integer calculations using Integer Roofline analysis.
• Integrated Roofline for integer data (PREVIEW) shows which exact memory layer is the bottleneck for each loop.
• Use the macOS* user interface to view and analyze data collected on Linux or Windows systems.
• Flow Graph Analyzer is a new rapid visual prototyping environment to interactively build, validate, and visualize algorithms.
• Reduce overhead for faster analysis results using selective profiling:
  • Decrease overhead for Memory Access Patterns and Dependencies analyses via limiting loop call count and analysis duration in project properties.
  • Selective profiling for Roofline, FLOPS and Trip Counts collections to decrease analysis scope and decrease overhead.
• Scale Roofs on the Roofline chart to arbitrary thread counts to see an application's practical performance limits.
• Compare several Roofline results on the same chart.
• New recommendation to optimize standard algorithms in C++ with Parallel STL.
• Supports new operating systems: Fedora 28, Red Hat Enterprise Linux 7.5, Microsoft Windows 10 build 17134, Ubuntu 18.04, SUSE Linux Enterprise Server (SLES) 12 SP3 and 15, macOS 10.11.x, 10.12.x, and 10.13.x.

Intel® Inspector

• Adds Persistence Inspector feature.
• Inspector debugger add-on now supports Microsoft Visual Studio* 2017.
• Supports new operating systems: Fedora 28, Red Hat Enterprise Linux 7.5, Microsoft Windows 10 build 17134, Ubuntu 18.04, SUSE Linux Enterprise Server* (SLES) 12 SP3and 15.

OpenCL™ Tools -- New Tools!

• Improved support for 8th Generation Intel® Core™ Processors (code-named Coffee Lake) on Windows and Linux operating systems.
• Offline compiler supports OpenCL™ Runtime 18.1 for Intel® Core™ Processors with added support of Intel® Advanced Vector Extensions 512 (Intel® AVX-512).
• Supports Intel® Streaming SIMD Extensions for Intel® FPGA Emulation Platform for OpenCL™.
• Adds support for additional operating systems and platforms, and compatible integration into more recent IDEs for flexibility.
• New features help speed development and improve performance for creating high-performance image and video processing pipelines.

Eclipse* IDE

• Updated Intel System Studio's IDE for Linux targets with Eclipse Oxygen 3 support.
• Provides sample projects for connecting to cloud services providers, including Microsoft Azure*, Amazon Web Services*, Google Cloud IoT Core* (new!), and IBM Bluemix.

Yocto Project* Compatible Application and Platform Project Development

An Eclipse plug-in enables users to create or import application projects and platform projects for Yocto Project* compatible targets running on Wind River Linux* LTS 17 (and above) or Yocto Project 2.6 (and above), using the Eclipse* IDE and the Yocto Project Toaster tool on Linux* development hosts.

Features

The Application Project feature enhances support for application development for embedded Linux devices, complete with multiple SDK support, debug flags enablement, customizable build properties, sample applications, target connections, run/debug launch configurations, and debugging. This feature includes the ability to directly use the Intel® C/C++ Compiler, together with a set of sample Intel C/C++ Compiler applications.

The Platform Project feature provides immediate out-of-box guidance for new users on how to create projects, and guided support for the most common first steps in creating and deploying a project, including integration of Intel System Studio agents.

This feature set provides out-of-box-experience guidance to help create and deploy embedded Linux applications and projects in the Intel System Studio graphical user interface (GUI), plus easy transition to and from the GUI and the command line. It also easily integrates with System Studio features.

This feature works with Yocto Project community of products and commercial partners.

Usage

• Configure and build an application
  • Import multiple cross-build SDKs, with each CPU variant immediately selectable.
  • Simple check box to enable debug versions of the application build.
  • Easy deploy and connection between the runtime target and the source-level debugger.
• Configure Platform Project file system via the web-based Toaster tool:
  • Full and guided access to the vast libraries of packages and features in Yocto Project* compatible universe.
  • Easily enables Intel System Studio agents and runtime Intel® C/C++ compiler support.
• Configure Platform Project kernel via the Kernel Configure view:
  • Ability to easily see and find hidden and/or disabled configuration items.
  • Provides easy reference calculations for dependencies, including reverse dependencies from the select and imply keywords.
  • Easy search and link tools.
  • Easy access to the parent KConfig files and their content.

Ubuntu* 18.04 LTS Linux* Support

• Supports Ubuntu* 18.04 LTS Linux* as a host development environment.

Intel® C++ Compiler 19.0

• -openmp-simd is set by default.
• -rcd option was deprecated.
• Changes mitigate speculative executive side channel and new -mindirect-branch option.
• Supports new C++ 17 features.

Intel® Data Analytics Acceleration Library (Intel® DAAL)

• Enables support for user-defined data modification procedure in CSV and ODBC data sources for implementing a wide range of feature extraction and transformation techniques on the user side.

Intel® Math Kernel Library (Intel® MKL)

• Adds automatic S/DGEMM JIT capability for small matrix sizes (m,n,k <=16) to improve S/DGEMM performance for Intel® Advanced Vector Extensions 2 (Intel® AVX2) and Intel® AVX-512 when compiling with MKL_DIRECT_CALL_JIT (threaded usage) or MKL_DIRECT_CALL_SEQ_JIT (sequential usage).
• Adds new functions to JIT (create) optimized S/DGEMM-like matrix multiply kernels for small matrix sizes (m,n,k <=16) for Intel® Advanced Vector Extensions 2 (Intel® AVX2) and Intel® Advanced Vector Extensions 512 (Intel® AVX-512), execute the optimized kernel created using matrices with matching dimensions, and to remove (destroy) the JIT kernel.
• Adds SparseQR and Extreme{EVD/SVD} functionality.
• Adds Multinominal Random Number Generators.

Intel® Integrated Performance Primitives (Intel® IPP)

• Adds new functions to support the ZFP floating-point data compression and decompression. Adds optimization patch files for ZFP 0.5.2 source to provide drop-in optimization with the Intel® IPP functions.
• Adds color conversion functions to convert RGB image to CIE Lab color model, or CIE Lab color model to RGB (ippiRGBToLab / ippiLabToRGB).
• Extends optimization for Intel® AVX-512 and Intel® AVX2 instruction sets in the ippiFilterRowPipeline and ippiMean_StdDev APIs.
• Improves LZ4 compression and decompression performance on data with high entropy.

Intel® Threading Building Blocks (Intel® TBB)

• Improves support for Flow Graph Analyzer and Intel VTune Amplifier in the task scheduler and generic parallel algorithms.
• Default device set for opencl_node now includes all the devices from the first available OpenCL™ platform.
• Adds lightweight policy for functional nodes in the flow graph. It indicates that the node body has little work and should, if possible, be executed immediately upon receiving a message, avoiding task scheduling overhead.

IoT connection tools: MRAA & UPM Libraries

       MRAA IO Communication Layer

• Supports new platforms: IEI Tank*, Ultra-96*, Rock960*, Hikey960*, and Dragonboard 820c*.
• Adds support for chardev GPIO devices with implementation on several boards.
• Adds string based I/O initialization for simpler integration with other frameworks.
• Adds GTest support for unit testing.
• Removes packaging and CI for obsolete devkits.
• Fixes multiple bugs and cmake improvements.

       UPM Sensor and Actuator Library

• Adds Cordova* bindings based on Java packages.
• Complete rework of the nmea_gps library giving more control over data acquisition and parsing.
• Adds GTest support for unit tests.
• Removes calls to deprecated MRAA fast GPIO functions.
• Improved SWIG type maps and gcc8 support.
• Supports new sensors: kx122, kxtj3, bh1792, and bh1749.

Intel® System Debugger

• Simplifies target and connection type configuration, and manage multiple simultaneous debug or trace connections to quicken system development.
• New Python* scripting interface provides the ability to configure and automate system trace runs on the command line.
• Hardware Support: 8th Generation Intel® Core™ U- and S-Series Processors, and Intel® Xeon® Processor D-1500 product family.

Intel® Debug Extensions for WinDbg*

• Intel® Processor Trace timestamps over direct connect interface (DCI) help triage timing and performance issues.

GNU* GDB and source

• Removed Intel® Debugger for Heterogeneous Compute.

2018

Update 2

Release Notes FreeBSD* Target Release Notes System Requirements

Overview:

• Intel® VTune™ Amplifier extends Input and Output analysis on Linux* to profile DPDK and SPDK IO APIs. Use this data to correlate CPU activity with the network data plane utilization, visualize PCIe bandwidth utilization per NIC, estimate UPI bandwidth on multi-socket systems, and identify bottlenecks. Application Performance Snapshot report generation time is improved, and adds an aggregated communication time column to the Rank-to-Rank data transfers report.
   
• Energy Analysis / Intel® SoC Watch collection metrics and group names are now more consistent across target operating systems.
   
• Intel® Debug Extensions for WinDbg* supports event-based breakpoints to debug ACPI Machine Language (AML) and can collect BSOD information with the get_bsod_info script.
   
• Includes the latest updates for many of the performance libraries, and analysis and debugger tools.

Detailed New Features List

Eclipse* IDE

• Upgrades JRE to July 17, 2018 CPU release to fix CVEs.
• Updates MRAA, UPM, and sensor support.
• Updates system sample builds to not report syntax errors for samples that compile successfully.
• Fixes previously reported issues.

Intel® C++ Compiler 18.0

• Changes mitigate speculative executive side channel and new -mindirect-branch option.
• More stable integration with Microsoft Visual Studio* 2017.
• Fixes previously reported issues.

Intel® Data Analytics Acceleration Library (Intel® DAAL)

• Resolves several issues.

Intel® Math Kernel Library (Intel® MKL)

• Improves BLAS performance on small sizes of multithreaded {S,D}SYRK and {C,Z}HERK for Intel® Advanced Vector Extensions 2 (Intel® AVX2) and Intel® Advanced Vector Extensions 512 (Intel® AVX-512).
• Adds LAPACK ?POTRF and ?GEQRF optimizations for Intel® AVX2 and Intel® AVX-512 instruction sets.
• Improves LAPACK performance of ?GESVD for very small square matrices (N<6) and inverse routines ?TRTRI, ?GETRI and ?POTRI.
• Improves SparseBLAS performance of SPARSE_OPTIMIZE, SPARSE_SV and SPARSE_SYPR routines for Intel® TBB threading.
• Enables optimizations for Intel® AVX-512 instruction set with support of Vector Neural Network Instructions via MKL_ENABLE_INSTRUCTIONS.

Intel® Integrated Performance Primitives (Intel® IPP)

• Improves LZ4 compression and decompression performance on data with high entropy.
• Resolves several issues and improves mitigation for security vulnerability CVE-2018-3617 in the Intel® IPP Cryptography libraries.

Intel® Threading Building Blocks

• Improves support for Flow Graph Analyzer and Intel® VTune™ Amplifier in the task scheduler and generic parallel algorithms.
• Adds lightweight policy for functional nodes in the flow graph to indicate the node body has little work and should, if possible, be executed immediately upon receiving a message, avoiding task scheduling overhead.

IoT connection tools: MRAA & UPM Libraries

Intel® VTune™ Amplifier

• Extends Input and Output analysis (formerly, Disk Input and Output analysis) on Linux* to profile DPDK and SPDK IO APIs. Use this data to correlate CPU activity with the network data plane utilization, visualize PCIe bandwidth utilization per NIC, estimate UPI bandwidth on multi-socket systems, and identify bottlenecks.
• Improves Application Performance Snapshot report generation time, allows reports to be generated in either text (*.txt) or comma-separated (*.csv) format, and adds an aggregated communication time column to the Rank-to-Rank data transfers report.
• Supports newer versions of several OSs.

Energy Analysis / Intel® SoC Watch

• Adds "--log " option for logging console output to a file.

       Intel SoC Watch for Linux* / Android*

• Separates HWP reporting into its own feature name to allow finer grain selection of which metrics to collect. Use hw-cpu-hwp to get this data now, rather than cpu-pstate.
• Features cpu-cstate and gfx-cstate are now group names. The individual feature names are now hw-cpu-cstate and hw-gfx-cstate to align these names across all operating systems.
• The -f hw-cpu-pstate feature is now collected by sampling so this metric can be supported on any system. Since this is now sampled data, there is some loss in precision of the CPU P-state residency and the report format changes.

       Intel® SoC Watch for Windows*

• Removes OS-based metrics that have comparable hardware metrics from the sys group to reduce post-processing time. The os-cpu-cstate, os-cpu-pstate, and os-gfx-cstate features should be listed explicitly when needed.
• Separates HWP reporting into its own feature name to allow finer grain selection of which metrics to collect. Use hw-cpu-hwp to get this data now, rather than hw-cpu-pstate.

Intel® Graphics Performance Analyzers

• Frame Analyzer for DirectX* 11 and DirectX 12 applications.
• Identify problematic textures within a draw call with new additions to 2 x 2 Texture Experiment. Replace a single input texture with a simple 2 x 2 texture or clamp to a selected MIP map level to see the performance delta and determine overall draw call impact.
• View and optimize the per-target, post-transformation geometry for a range of selected draw calls in DirectX 11 frames.

Intel® Inspector

• Inspector debugger add-on now supports Microsoft* Visual Studio* 2017.
• New OS support: SUSE* Linux* Enterprise Server* 12 SP3, Microsoft* Windows* 10 build 17134
• Bug fixes

Intel® System Debugger

Intel® Debug Extensions for WinDbg*

• Adds support for event-based breakpoints to debug ACPI Machine Language (AML).
• Adds ability to collect BSOD information with the get_bsod_info script.

Update 1

Release Notes FreeBSD* Target Release Notes System Requirements

Overview:

• Move from prototype to product easier with new capabilities that seamlessly import applications from Arduino Create* to Intel System Studio. Take advantage of System Studio’s advanced analyzers and debug tools for advanced system development. More tools and libraries also now support the Up Squared* Grove* IoT Development Kit. Learn more.
   
• Java* support – Developers can now create, build, run and edit native Java applications using Intel System Studio. Through the Project Creation Wizard, use the cloud connectors and access Intel IoT sensor libraries. Java examples can also be enabled for the Up Squared* Grove* IoT Development Kit.
   
• Easier access to code samples, automated configuration – Using code samples is now easier using the new Project Creation Wizard. All required configuration options are automatically set when sample projects are created.
   
• Connect with various cloud service providers’ APIs more simply via the cloud connector API explorer.
   
• Includes the latest updates for many of the performance libraries, and analysis and debugger tools.

Detailed New Features List

Eclipse* IDE

• Eclipse IDE on Linux* Ubuntu* 16.04.4 LTS now depends on GTK3, Ubuntu 17 and 18 will continue to use GTK2.
• Platform Manager now performs verbose Docker* image builds. Project Creation Wizards and Sensor Explorer have been streamlined for better user experience.
• Supports development of Wind River Linux* LTS* 17 applications. This is only supported on a Linux* host and does not support creation of an LTS 17 kernel image.

Intel® C++ Compiler 18.0

• More stable integration with Microsoft Visual Studio* 2017
• Fixes previously reported issues

Intel® Data Analytics Acceleration Library (Intel® DAAL)

• Host application interface is added to Intel® DAAL, which enables algorithm-level computation cancelling by user-defined callback. This interface is available in Decision Forest and Gradient Boosting Trees algorithms. New example code is provided.
• New technical preview for experimental Intel DAAL and its extension library
  • Introduced distributed k-Nearest Neighbors classifiers for both training and prediction. Included new sample that demonstrates how to use this algorithm with Intel® MPI Library.
  • Developed experimental extension library on top of existing pyDAAL package that provides an easy to use API for Intel DAAL neural networks. The extension library allows users to configure and train neural network models in a few lines of code, and the ability to use existing TensorFlow* and Caffe* models on inference stage.
• Gradient Boosting Trees training algorithm was extended with inexact splits calculation mode. It is applied to continuous features that are bucketed into discrete bins, and the possible splits are restricted by the buckets borders.
• Intel® Threading Building Blocks (Intel® TBB) dependency is removed in library sequential mode.

Intel® Math Kernel Library (Intel® MKL)

Intel® Integrated Performance Primitives (Intel® IPP)

IoT connection tools: MRAA & UPM Libraries

MRAA IO Communication Layer

• New APIs for sysfs onboard LED control using the gpio-leds driver
• Restructured and cleaned-up basic examples

UPM Sensor and Actuator Library

• Extended LED library to support the new MRAA gpio-leds APIs
• Cleaned-up doxygen tags in headers and class names in JSON library files to facilitate integration with the Sensor Explorer

Intel® VTune™ Amplifier

• New CPU/FPGA interaction analysis (Technical Preview) to assess the balance between the CPU and FPGA on systems with a discrete Intel® Arria® 10 FPGA running OpenCL™ applications
• New Graphics Rendering analysis (Technical Preview) for CPU/GPU utilization of your code running on the Xen* virtualization platform installed on a remote embedded target
• Support for the sampling command-line analysis on remote QNX* embedded systems via ethernet connection

Energy Analysis / Intel® SoC Watch

       Intel® SoC Watch for Windows*

• Add support for Intel platform code named Gemini Lake
• Resolves several issues

       Intel® SoC Watch for Linux* / Android*

• Add support for Intel® platform codenamed Gemini Lake.
• New feature group “sstate” added: Measures both operating system (Sx) and hardware (S0ix) platform sleep states on platforms that measure both.

Intel® Inspector

• Deadlocks detection on std::shared_mutex (C++17 standard)
• New OS support - Fedora Core* 27, Ubuntu* 17.10, Microsoft Windows* 10 RS3
• Bug fixes

Intel® System Debugger

Intel® Debug Extensions for WinDbg*

• Support for event-based breakpoints to debug ACPI Machine Language (AML)
• Added feature to collect BSOD information with the get_bsod_info script

GNU* GDB and source

• GDB Server is supported on Wind River Linux* LTS 17.
• The GDB Server binaries from the WindriverLinux9 directory can also be used for Wind River Linux* LTS 17.

Initial Release

Release Notes System Requirements

Overview:

• New libraries and code samples help shorten the development cycle. By using the Intel® Data Analytics Acceleration Library, developers can speed edge analytics processing and machine learning.
• New IoT connection tools including advanced cloud connectors and access to 400+ sensors.
• Support for the latest Intel® processors - utilize Intel® AVX-512 instructions to optimize system and code performance.¹
• Debug capabilities and enhanced workflows that ease system validation for target devices, automate tracing, ensure reliable edge-to-cloud data exchange, and more.
• New ability to customize your software download - get only the tools you need.

Detailed New Features List

Eclipse* IDE for Intel System Studio 2018

• Created an Intel version of the Eclipse* IDE for Intel System Studio 2018
• Created modular Eclipse IDE structure for contribution to the Intel System Studio product
• Integrated Intel System Studio for IoT Edition into Intel System Studio 2018
• Improved remote Linux* OS target support 
  • Added Eclipse Target Communication Framework support for target connection
  • Added basic Makefile support
• Added wizards for Intel® C++ Compiler integration
  • Added local compiler integration for Linux hosts
  • Added cross-compilation integration with support for Linux and Android* OS targets
• Improved general user experience
  • Custom perspectives
  • Implemented wizards focusing on Intel System Studio use cases
  • Disabled unsupported wizards

Intel® C++ Compiler 18.0

• Control-Flow Enforcement Technology (CET) support
• New option -Qimf-use-svml to enforce short vector math library (SVML)
• Compile-time dispatching for SVML calls
• All -o* options replaced with -qo* options
• Support of hardware based Profile Guided Optimization (PGO)
• Features from OpenMP* TR4 Version 5.0 Preview 1
• Support for more new features in OpenMP* 4.0 or later
• New C++17 features supported
• Support for the atomic keyword introduced in C++11
• New option –qopt-zmm-usage that defines the level of ZMM registers usage

Intel® Math Kernel Library 2018 (Intel® MKL)

• BLAS Features
  • Introduced 'compact GEMM' and 'compact TRSM' functions to work on groups of matrices and added service functions to support the new format.
  • Introduced optimized integer matrix-matrix multiplication routine to work with quantized matrices for all architectures.
• BLAS Optimizations
  • Optimized GEMM_S8U8S32 and GEMM_S16S16S32 for Intel® Advanced Vector Extensions 2 (Intel® AVX-2), and Intel® AVX-512 with support of AVX512_4FMAPS and AVX512_4VNNIW instruction groups.
• Deep Neural Network
  • Added support for non-square pooling kernels.
  • Optimized conversions between plain (nchw, nhwc) and internal data layouts.
• LAPACK
  • Added improvements and optimizations for small matrices (N<16).
  • Added ?gesvd, ?geqr/?gemqr, ?gelq/?gemlq optimizations for tall-and-skinny and short-and-wide matrices.
  • Added optimizations for ?pbtrsroutine.
  • Added optimizations for ?potrf routine for Intel® Threading Building Blocks (Intel® TBB) layer.
  • Added optimizations for CS decomposition routines:?dorcsd and?orcsd2by1.
  • Introduced factorization and solve routines based on Aasen's algorithm:?sytrf_aa/?hetrf_aa, ?sytrs_aa/?hetrs_aa.
  • Introduced new (faster)_rk routines for symmetric indefinite (or Hermitian indefinite) factorization with bounded Bunch-Kaufman (rook) pivoting algorithm.
• ScaLAPACK
  • Added optimizations (2-stage band reduction) for p?syevr/p?heevr routines for JOBZ=’N’ (eigenvalues only) case.
• FFT
  • Introduced Verbose support for FFT domain, which enables users to capture the FFT descriptor information for Intel® MKL.
  • Improved performance for 2D real-to-complex and complex-to-real matrix multiplication for Intel® Xeon® processors supporting Intel AVX-512.
  • Improved performance for 3D complex-to-complex for Intel Xeon processors supporting Intel AVX-512.
• Intel Optimized High Performance Conjugate Gradient Benchmark
  • New version of benchmark with Intel MKL API
• Sparse BLAS
  • Introduced Symmetric Gauss-Zeidel preconditioner with ddot calculation of resulted and initial arrays.
  • Sparse Matvec routine with ddot calculation of resulted and initial arrays.
  • Sparse Syrk routine with both OpenMP and Intel® TBB support.
  • Improved performance of Sparse MM and MV functionality for Intel AVX-512 instruction set.
• Direct Sparse Solver for Cluster
  • Add support of transpose solver
• Vector Mathematics
  • Added 24 functions including optimizations for processors based on Intel AVX-512.
• Data Fitting
  • Cubic spline-based interpolation in ILP64 interface was optimized up to 8x times on Intel Xeon processors supporting Intel AVX-512.

Intel® Data Analytics Acceleration Library(Intel® DAAL)

• Introduced API modifications to streamline library usage and enable consistency across functionality.
• Introduced support for Decision Tree for both classification and regression. The feature includes calculation of Gini index and Information Gain for classification, and mean squared error (MSE) for regression split criteria, and Reduced Error Pruning.
• Introduced support for Decision Forest for both classification and regression. The feature includes calculation of Gini index for classification, variance for regression split criteria, generalization error, and variable importance measures such as Mean Decrease Impurity and Mean Decrease Accuracy.
• Introduced support for varying learning rate in the Stochastic Gradient Descent algorithm for neural network training.
• Introduced support for filtering in the Data Source including loading selected features/columns from CSV data source and binary representation of the categorical features
• Extended Neural Network layers with Element Wise Add layer.
• Introduced new samples that allow easy integration of the library with Spark* MLlib
• Introduced service method for enabling thread pinning; performance improvements in various algorithms on Intel Xeon processors supporting Intel AVX-512.

Intel® Integrated Performance Primitives 2018 (Intel® IPP)

• Optimized functions for LZ4 data compression and decompression, a fast compression algorithm suitable for applications where speed is key - especially in communication channels.
• Optimized functions for GraphicsMagick*, a popular image processing toolbox, so customers using this function can achieve improved performance using drop-in optimization with Intel® IPP functions.
• Removed the cryptography code dependency on the main package.
• Extended support of platform-aware APIs, which automatically detects whether image vectors and length are 32-bit or 64-bit, provides 64-bit parameters for image dimensions and vector length, and abstracts this away from the users.

Intel® Threading Building Blocks 2018 (Intel® TBB)

• this_task_arena::isolate() function is now a fully supported feature. Also, this_task_arena::isolate() function and task_arena::execute() methods were extended to pass on the value returned by the executed functor (this feature requires C++11). The task_arena::enqueue() and task_group::run() methods extended to accept move-only functors.
• Added support for Android* NDK r15, r15b.
• Added support for Universal Windows Platform*.

IoT connection tools: MRAA & UPM Libraries

• Includes more than 400 sensor and actuator libraries, with a built-in GUI for exploring the repository
• Support for these libraries included for Ubuntu*, Wind River Linux*, and Wind River Pulsar*
• Additional samples included which show how to leverage MRAA and UPM in combination with various cloud services.

Intel® VTune™ Amplifier 2018

• Easier Analysis of Remote Linux Systems
  • Automated install of Intel® Vtune™ Amplifier collectors on a remote Linux target.
• Enhanced Python* Profiling
  • Locks and Waits analysis tunes threaded performance of mixed Python* and native code.
  • Preview: Memory consumption analysis. Python, C, C++.
• Optimize Private Cloud-Based Applications
  • Profile inside Docker & Mesos containers.
  • Attach to running Java services and daemons.
• Media Developers: GPU In-kernel Profiling
  • Analyze GPU kernel execution to find memory latency or inefficient kernel algorithms.
• Easier Threading Optimization of Applications Using Intel TBB
  • Advanced threading analysis extends classification of high overhead and spin time.
• Latest Processors
  • New Intel® processors including Intel Xeon Scalable processor.
• Cross OS Analysis for All Supported OSes
  • Download other OSes as needed. e.g., collect data on Linux, then analyze it on Windows* or macOS*.

Energy Analysis/Intel® SoC Watch

• Added Eclipse* Plug-in for Energy analysis [Preview]

Intel® Inspector 2018

• Support for C++17 std::shared_mutex and Windows SRW Locks, that enable threading error analysis for applications with read/write synchronization primitives.
• Support for cross-OS analysis to all license types. The installation packages for additional operating systems can be downloaded from registrationcenter.intel.com.
• Microsoft Visual Studio 2017* integration and support.

Intel® Graphics Performance Analyzers

• Multi-Frame Analyzer Feature Pack 1
• Trace Analyzer PA Replacement
• 8th Gen Intel® Core™ Processor(formerly Kaby Lake Refresh) Windows 10 support
• Windows Redstone 3 support

Intel® System Debugger 2018

• Added new method for connecting to target systems, called Target Connection Agent.
• Support for Intel Atom® Processor C3xxx target added for both Windows and Linux hosts.
• Support for Intel® Xeon® Scalable Processor / Intel® C620 Series chipset target added for Windows host.
• Support for 8th generation Intel® Core™ Processor / Intel® 100 Series Chipset added for Windows host.
• Support for “8th generation Intel® Core™ processor / Intel®  Z370 Series Chipset target added for Windows host.

Intel® Debug Extensions for WinDbg*

• WinDbg* supports Windows Driver Kit (WDK) version 1703. Added support for a new eXDI callback (DBGENG_EXDI_IOCTL_V3_GET_NT_BASE_ADDRESS_VALUE) to locate windows key structure KdVersionBlock.
• Extended Intel® Debug Extensions for WinDbg* for Intel® Processor Trace plug-in to support Windows public symbol information.
• Extended Intel Debug Extensions for WinDbg* for Intel Processor Trace plug-in to support ring 3 tracing.
• Extended Intel Debug Extensions for WinDbg* for Intel Processor Trace plug-in to support decoding Intel Processor Trace data from crash dump.

GNU* GDB and source

• Added visualizer for PKeys hardware register and GS_base and FS_base system registers in Linux.
• Added Python* call backs for Intel® Processor Trace.

2017

Update 3

Linux* Host Release Notes Windows* Host Release Notes Windows* Target Release Notes FreeBSD* Target Release Notes

Overview:

• Advanced debug, trace and analysis features and reports help developers find and fix issues and optimize code.
• Compilers and libraries provide innovative features and improved performance for high-demand processing, including image, machine learning, storage, communications, transportation usages and more.
• Multiple usability and workflow enhancements, and new tutorials make development easier.
• The 2017 version supports the latest Intel processor-based platforms and offers multiple editions to meet developer needs.
   

Detailed New Features List

Support for the Latest Intel Processor-Based Embedded, Mobile, and IoT Platforms

• Intel Atom® processors Z5xx, N4xx, N5xx, D5xx, E6xx, N2xxx, D2xxx, Z2xxx, Z3xxx, E3xxx, C2xxx, CE4xxx, CE53xx and the Intel® Puma™ 6 Media Gateway
• Intel® Pentium® Processor N4200, Intel® Celeron® Processor N3350, Intel Atom® Processors x7-E3950, x5-3940, x3-3930 (Broxton Apollo Lake)
• Intel® Edison development platform
• 2nd , 3rd , 4th, 5th or 6th generation Intel® Core™ processor
• Intel® Xeon® processors based on 2nd, 3rd 4 th or 5th generation Intel® Core™ architecture
• 5th generation Intel® Core™ M processor

Eclipse* IDE

• Complete IDE with simplified workflow
• Added more task-based tutorials
• Added option to download files for later installation

Intel® C++ Compiler 17.0

• Fixes to reported problems.

Intel® Integrated Performance Primitives 2017

• Fixed some known problems in Intel® IPP Cryptography functions. Visit Intel® IPP 2017 bug fixes for the bug fix list.
• Added support for Microsoft Visual Studio* 2017 on Windows*.

Intel® Math Kernel Library 2017

• BLAS
  • Improved performance for ?GEMM for medium problem sizes on Intel® Xeon® Processor supporting Intel® Advanced Vector Extensions 512 (Intel® AVX-512) (codename Skylake Server)
  • Improved performance for SGEMM/DGEMM for small problem sizes on Intel® Xeon® Processor supporting Intel® Advanced Vector Extensions 512 (Intel® AVX-512) (codename Skylake Server)
  • Improved ?GEMM_BATCH performance on all architectures
  • Improved performance for SSYMV/DSYMV on Intel® Advanced Vector Extensions 2 (Intel® AVX2) and later architectures
  • Improved performance for general BLAS functions on the 32-bit Intel® Advanced Vector Extensions 512 (Intel® AVX-512) architecture
  • Fixed ?AXPBY to propagate NaNs in the y vector when beta = 0 on 64-bit Intel® Advanced Vector Extensions 2 (Intel® AVX2) and later architectures
• High Performance Conjugate Gradients (HPCG)
  • Add support of Intel® Xeon® Processor supporting Intel® Advanced Vector Extensions 512 (Intel® AVX-512) (codename Skylake Server)
• Deep Neural Networks
  • Average pooling has an option to include padding into mean values computation.
• LAPACK
  • Optimized ?GELQ and ?GEMLQ performance for short-and-wide matrices.
  • Optimized performance of ?ORCSD2BY1 and ?DORCSD routines.
  • Fixed LU performance degradation for medium sizes on 6 threads.

Intel® Threading Building Blocks 2017

• Added support for Android* NDK r14.
• Added a blocking terminate extension to the task_scheduler_init class that allows an object to wait for termination of worker threads.
• Fixed compilation and testing issues with MinGW (GCC 6).
• Fixed compilation with /std:c++latest option of Microsoft Visual Studio* 2017 (https://github.com/01org/tbb/issues/13).

Intel® VTune™ Amplifier 2017 for Systems

• Application Performance Snapshot (Preview) provides a quick look at your application performance and helps you understand where your application will benefit from tuning. The revised tool shows metrics on MPI parallelism (Linux* only), OpenMP* parallelism, memory access, FPU utilization, and I/O efficiency with recommendations on further indepth analysis.
• Improved insight into parallelism inefficiencies for applications using Intel® Threading Building Blocks (Intel® TBB) with extended classification of high Overhead and Spin time.
• Automated installation of the VTune Amplifier collectors on a remote Linux* target system. This feature is helpful if you profile a target on a shared resource without VTune Amplifier installed or on an embedded platform for Microsoft Visual Studio* 2017.

SoC Watch

• Initial support for Intel Atom® E39xx platforms and 7th generation Intel® Core™ processors.
• pkg-temp on Intel Atom® E39xx platforms: Measure the thermal status of multiple domains in the package.
• dram-bw on Intel Atom® E39xx platforms: Reports total DDR bandwidth with memory slice breakdown.

Intel® Inspector 2017

• Correctness improvements of uninitialized memory reads detection algorithm.
• Fixes for reported bugs.

Intel® System Debugger 2017

• System Debug
  • Providing information to the users about breakpoint re-application after target reset.
  • Updated Intel® DFx Abstraction Layer (Intel® DAL).
  • I/O port based reset shows now an information which port writes were executed.
  • Improved System Table discovery for LoadDXEModules function.
  • The Target Access Module hang detection timeout can now be configured in the options dialog. Especially for server platforms with a large number of threads increasing the timeout helps to identify real target access module hangs.
  • Intel® Processor Trace Plugin added support for multiple configuration tables and regions.
  • Providing an appropriate message if a target access call is taking a long time instead of trying to abort the call.
  • Fixed crash during usage of the LoadPEIM function if PEI modules were built in release mode.
  • I/O port based reset does not use to default reset method anymore but do a port based reset.
• System Trace
  • Add support for the new OpenIPC API.
  • Timeline View: Add Context Menu for better accessibility. The menu contains items available in the editor toolbar.
  • Timeline View: The timeline view height axis is reduced to save vertical space. Keyboard shortcuts are added to all the commands available in the menu. A new command – “Suspended Details” – is added.
  • Message Router: Watchdog for Intel Trace Hub Server has been added. Heartbeat mechanism has been added between UI and the native side. This was done following RFC6455 when the client sends PING message (0x09) and the server responds with PONG message (0x0A).
  • Fixed missing mnemonics on XML serialization, which caused a delay in decode of certain formats.
  • Fixed an issue related to ignoring the user extension destination. The collateral installer was ignoring the destination folder setting from the manifest.

2016

Update 4

Linux* Host Release Notes Windows* Target Release Notes

Intel® C++ Compiler

• Update to version 16.0.4
• Support for Microsoft* Visual Studio* 2015 Update 3

Intel® Math Kernel Library (Intel® MKL)

• Update to version 11.3 Update 4
• BLAS
  • Introduced new packed matrix multiplication interfaces (?gemm_alloc, ?gemm_pack ,?gemm_compute, ?gemm_free) for single and double precisions.
  • Improved performance over standard S/DGEMM on Intel® Xeon® processor E5-xxxx v3 and later processors.
• LAPACK
  • Improved LU factorization, solve, and inverse (?GETR?) performance for very small sizes (<16).
  • Improved General Eigensolver (?GEEV and ?GEEVD) performance for the case when eigenvectors are needed.
  • Added Intel® Threading Building Blocks (Intel® TBB) parallelism for ?ORGQR/?UNGQR.

Intel® Integrated Performance Primitives (Intel® IPP)

• Update to version 9.0 Update 4

Intel® Threading Building Blocks (Intel® TBB)

• Update to version 4.4 Update 6
• For 64-bit platforms, quadrupled the worst-case limit on the amount of memory the Intel® TBB allocator can handle.
• Fixed a memory corruption in the memory allocator when it meets internal limits.
• Fixed the memory allocator on 64-bit platforms to align memory to 16 bytes by default for all allocations bigger than 8 bytes.
• Fixed parallel_scan to provide correct result if the initial value of an accumulator is not the operation identity value.
• As a workaround for crashes in the Intel® TBB library compiled with GCC 6, added -flifetime-dse=1 to compilation options on Linux* OS

Intel® System Debugger

• Update to version 2016 Update 4
• Fix for AET decoder crash after multiple start/stop cycles
• System Trace
  • Architectural Event Traces (AET) support added
  • CSME verbosity can be set to "Verbose" or "Normal" in the configuration editor.
  • Eclipse* Neon (4.6) supported
  • New buttons for de-/selecting all traces sources in the Event Distribution View (EDV).

Intel® Graphics Performance Analyzers (Intel® GPA)

• Update to version 2016 R2

Update 3

FreeBSD* Target Release Notes Linux* Host Release Notes Windows* Target Release Notes

Intel® C++ Compiler

• Annotated source listing: annotates source files with compiler optimization reports. The listing format may be specified as either text or html.
• New attribute, pragma, and compiler options for code alignment
• Additional C++14 features supported
• Additional C11 features supported

Intel® Math Kernel Library (Intel® MKL)

• Introducing Deep Neural Networks (DNN) primitives including convolution, normalization, activation and pooling functions intended to accelerate convolutional neural networks (CNNs) and Deep neural networks (DNNs) on Intel® architecture
• Removed SP2DP interface library
• Removed pre-compiled BLACS library for MPICH v1; MPICH users can still build the BLACS library with MPICH support via Intel MKL MPI wrappers
• Sparse BLAS
  • Improved performance of parallel BSRMV functionality for processor supporting Intel® Advanced Vector Extensions 2 (Intel® AVX2) instruction set.
• Intel MKL PARDISO
  • Added support for mkl_progress in Parallel Direct Sparse Solver for Clusters.
• DFT
  • Improved performance of batched 1D FFT with large batch size on processor supporting Intel® Advanced Vector Extensions (Intel® AVX), Intel® Advanced Vector Extensions 2 (Intel® AVX2) and Intel® Advanced Vector Extensions 512 (Intel® AVX-512) instruction sets.
• Data Fitting
  • Introduced 2 new storage formats for interpolation results (DF_MATRIX_STORAGE_SITES_FUNCS_DERS, DF_MATRIX_STORAGE_SITES_DERS_FUNCS).

Intel® Integrated Performance Primitives (Intel® IPP)

• Added new APIs (Intel® IPP 64x functions) to support 64-bit data length in the image and signal processing domains:
  • This release provides the 64x functions for memory allocation, image addition, subtraction, multiplication, division, resizing, and filtering operations.
  • The Intel® IPP 64x functions are implemented as wrappers over Intel® IPP functions operating on 32-bit sizes by using tiling and multithreading. The 64x APIs support external threading for Intel® IPP functions, and are provided in the form of source and pre-built binaries.
• Added integration wrappers for some image processing and computer vision functions. The wrappers provide the easy-to-use C and C++ APIs for Intel® IPP functions, and they are available as a separate download in the form of source and pre-built binaries.
• Performance and Optimization:
  • Extended optimization for Intel® Advanced Vector Extensions 512 (Intel® AVX-512) instruction set on Intel® Many Integrated Core Architecture (Intel® MIC Architecture). Please see the Intel® IPP Functions Optimized for Intel® AVX-512 article for more information.
  • Extended optimization for Intel® AVX-512 instruction set on Intel® Xeon® processors.
  • Extended optimization for Intel® Advanced Vector Extensions 2 (Intel® AVX2) instruction set on the 6th Generation Intel® Core™ processors. Please see the Intel® IPP Functions Optimized for Intel® AVX2 article for more information.
  • Extended optimization for Intel® Streaming SIMD Extensions 4.2 (Intel® SSE4.2) instruction set on Intel Atom® processors.
• Signal Processing:
  • Added the ippsIIRIIR functions that perform zero-phase digital IIR filtering.
  • Added 64-bit data length support to the ippsSortRadixAscend functions.
• Image Processing:
  • Added the ippiScaleC functions to support image data scaling and shifting for different data types.
• Data Compression:
  • Added the patch files for the zlib compression and decompression functions. The patches provide drop-in optimization with Intel® IPP functions, and support zlib version 1.2.5.3, 1.2.6.1, 1.2.7.3 and 1.2.8.
• Removed the tutorial from the installation package; its sample code and documentation are now provided online (https://software.intel.com/en-us/product-code-samples).
  • Threading Notes: Though Intel® IPP threaded libraries are not installed by default, these threaded libraries are available by the custom installation, so the code written with these libraries will still work as before. However, the multi-threaded libraries are deprecated and moving to external threading is recommended.

Intel® Threading Building Blocks (Intel® TBB)

• Removed a few cases of excessive user data copying in the flow graph.
• Improved robustness of concurrent_bounded_queue::abort() in case of simultaneous push and pop operations.
• Modified parallel_sort to not require a default constructor for values and to use iter_swap() for value swapping.
• Added support for creating or initializing a task_arena instance that is connected to the arena currently used by the thread.
• Added template class opencl_node to the flow graph API. It allows a flow graph to offload computations to OpenCL™ devices.
• Extended join_node to use type-specified message keys. It simplifies the API of the node by obtaining message keys via functions associated with the message type (instead of node ports).
• Added static_partitioner that minimizes overhead of parallel_for and parallel_reduce for well-balanced workloads.
• Improved template class async_node in the flow graph API to support user settable concurrency limits.
• Class global_control supports the value of 1 for max_allowed_parallelism.
• Added tbb::flow::async_msg, a special message type to support communications between the flow graph and external asynchronous activities.
• async_node modified to support use with C++03 compilers
• Fixed a bug in dynamic memory allocation replacement for Windows* OS.
• Fixed excessive memory consumption on Linux* OS caused by enabling zero-copy realloc.

Intel® System Debugger

• Support for Eclipse* 4.5 (Mars.2) for the trace viewer. The package is also included in the Intel® System Studio installation package for optional installation.
• Support for debug format Dwarf4
• SMM support for Intel® Core™ based processors debugging.
• A new EFI script and three buttons are added for loading PEI/DXE modules easily in System Debug.

Intel® VTune™ Amplifier for Systems

• Support for the next generation Intel® Xeon® Processor E5 v4 Family (formerly codenamed "Broadwell-EP")
• Detection of the OpenCL™ 2.0 Shared Virtual Memory (SVM) usage types per kernel instance
• Driverless event-based sampling collection for uncore events enabled for the Memory Access analysis.
• Support for the Microsoft* Visual Studio* 2015 Update 2
• Disk Input and Output analysis that monitors utilization of the disk subsystem, CPU and processor buses, helps identify long latency of I/O requests and imbalance between I/O and compute operations
• GPU Hotspots analysis targeted for GPU-bound applications and providing options to analyze execution of OpenCL™ kernels and Intel Media SDK tasks
• Basic Hotspots analysis extended to support Python* applications running via the Launch Application or Attach to Process modes.

Intel® Energy Profiler for Windows

• Update to version v1.14.1
• Extended collection start time information to include microseconds to better enable correlation with event trace logs.
• Corrected reporting of Gfx P-states on Intel® 6th Generation Core™ (formerly code-named “Skylake”) platform.

Intel® Graphics Performance Analyzers (Intel® GPA)

• Intel GPA now provides alpha-level support for DirectX* 12 application profiling. This version has limited profiling and debug capabilities and might work unstable on some workloads. You can find more details regarding the supported features below.
  • Graphics Frame Analyzer provides detailed GPU hardware metrics for Intel® graphics. For third-party GPUs, GPU Duration and graphics pipeline statistics metrics are available.
  • DirectX states, Geometry, Shader code, Static and dynamic textures, Render targets resources are available for frame-based analysis in Graphics Frame Analyzer.
  • Simple Pixel Shader, Disable Erg(s) performance experiments, Highlighting and Disable draw calls visual experiments are available in Graphics Frame Analyzer
  • Time-based GPU metrics for Intel graphics, CPU metrics, Media and Power metrics in System Analyzer.
  • System Analyzer HUD includes support for hotkeys, the same set of metrics as in System Analyzer, messages and settings.
  • Note: To capture DirectX 12 application frames, enable the Force DirectX12 injection option in the Graphics Monitor Preferences dialog box.
  • Note: System memory consumption is expected to be high in this release at both time of capture and during playback. Needed memory is related to workload and frame complexity and varies greatly. 8GB is minimum, 16GB is recommended, with some workloads requiring more.
• New Features for Analyzing OpenGL/OpenGL ES* Applications
  • Enabled support for GPU hardware metrics in System Analyzer and Graphics Frame Analyzer on the 6th Generation Intel® Core™ Processors for Ubuntu* targets.
  • Several OpenGL API calls (e.g. glTexImage2D, glReadPixels, glCopyTexImage2D, etc.) are now represented as ergs in Graphics Frame Analyzer, which allows measuring GPU metrics for them and see the used input and output.
• Implemented Resource History in Graphics Frame Analyzer. When you select a particular texture or program in the Resource viewer, colored markers appear in the bar chart, indicating the ergs where these resources are used. The color of these markers corresponds to the type of the resource: input, execution, or output.

Update 2

FreeBSD* Target Release Notes Linux* Host Release Notes Windows* Target Release Notes

Intel® C++ Compiler

• The Short Vector Random Number Generator (SVRNG) library provides intrinsics for the IA-32 and Intel® 64 architecture running on supported operating systems. The SVRNG library partially covers both standard C++ and the random number generation functionality of the Intel® Math Kernel Library (Intel® MKL). Complete documentation may be found in the Intel® C++ Compiler 16.0 User and Reference Guide.
• Intel® SIMD Data Layout Templates (Intel® SDLT)
  • Intel® SDLT is a library that helps you leverage SIMD hardware and compilers without having to be a SIMD vectorization expert.
  • Intel® SDLT can be used with any compiler supporting ISO C++11, Intel® Cilk™ Plus SIMD extensions, and #pragma ivdep
  • Intel® SIMD Data Layout Templates
• New C++14 and C11 features supported

Intel® Math Kernel Library (Intel® MKL)

• Introduced mkl_finalize function to facilitate usage models when Intel MKL dynamic libraries or third party dynamic libraries are linked with Intel MKL statically are loaded and unloaded explicitly.
• Introduced sorting algorithm
• Performance improvements for BLAS, LAPACK, ScaLAPACK, Sparse BLAS
• Several new features for Intel MKL PARDISO
• Added Intel® TBB threading support for all and OpenMP* for some BLAS level-1 functions.

Intel® Integrated Performance Primitives (Intel® IPP)

• Image Processing:
  • Added the contiguous volume format (C1V) support to the following 3D data processing functions: ipprWarpAffine, ipprRemap, and ipprFilter.
  • Added the ippiFilterBorderSetMode function to support high accuracy rounding mode in ippiFilterBorder.
  • Added the ippiCopyMirrorBorder function for copying the image values by adding the mirror border pixels.
  • Added mirror border support to the following filtering functions: ippiFilterBilateral, ippiFilterBoxBorder, ippiFilterBorder, ippiFilterSobel, and ippiFilterScharr.
  • Kernel coefficients in the ippiFilterBorder image filtering functions are used in direct order, which is different from the ippiFilter functions in the previous releases.
• Computer Vision:
  • Added 32-bit floating point input data support to the ippiSegmentWatershed function.
  • Added mirror border support to the following filtering functions: ippiFilterGaussianBorder, ippiFilterLaplacianBorder, ippiMinEigenVal, ippiHarrisCorner, ippiPyramidLayerDown, and ippiPyramidLayerUp.
• Signal Processing:
  • Added the ippsThreshold_LTAbsVal function, which uses the vector absolute value.
  • Added the ippsIIRIIR64f functions to perform zero-phase digital IIR filtering.
• The multi-threaded libraries only depend on the OpenMP* libraries; their dependencies on the other Intel® Compiler runtime libraries were removed.

Intel® System Debugger

• Unified installer now for all components of the Intel® System Debugger (for system debug, system trace and WinDbg* extension)
• Support for Eclipse* 4.4 (Luna) integration with Intel® Trace Viewer
• New "Trace Profiles" feature for System Trace Viewer to configure the destination for streaming mode for:
  • BIOS Reserverd Trace Memory
  • Intel® Trace Hub Memory
  • Streaming to DCI-Closed Chassis Adapter (BSSB CCA)
• Tracing to memory support (Intel® Trace Hub or system DRAM memory) for 6th Gen Intel® Core™ processors (PCH) via Intel® XDP3 JTAG probe.
• Various stability bug fixes in Trace Viewer: Handling of decoder-instance-parameters. Crash on stop capture. Errors resulting from renaming capture files. Fix for persistent page up/down navigation. Decoding linked files containing spaces in path. Sporadic Eclipse error when switching target.
• Trace Viewer improvements: Event distribution viewer. New progress bar when stopping a trace to memory. Rules are saved now in Eclipse workspace and restored during Eclipse restart. Improved memory download with wrapping enabled.
• Debugging support for Intel® Xeon® Processor D-1500 Product Family on the Grangeville platform.
• System Debugger improvements: Export memory window to text file.

Intel® VTune™ Amplifier for Systems

• Support for the ITT Counters API used to observe user-defined global characteristic counters that are unknown to the VTune Amplifier.
• Support for the Load Module API used to analyze code that is loaded in an alternate location that is not accessible by the VTune Amplifier.
• Option to limit the collected data size by setting a timer to save tracing data only for the specified last seconds of the data collection added for hardware event-based sampling analysis types.
• New Arbitrary Targets group added to create command line configurations to be launched from a different host. This option is especially useful for microarchitecture analysis since it provides easy access to the hardware events available on a platform you choose for configuration.
• Source/Assembly analysis available for OpenCL™ kernels (with no metrics data).
• SGX Hotspots analysis support for identifying hotspots inside security enclaves for systems with the Intel Software Guard Extensions (Intel SGX) feature enabled.
• Metric-based navigation between call stack types replacing the former Data of Interest selection.
• Updated filter bar options, including the selection of a filtering metric used to calculate the contribution of the selected program unit (module, thread, and so on).
• DRAM Bandwidth overtime and histogram data is scaled according to the maximum achievable DRAM bandwidth.

Intel® Energy Profiler for Windows

• Added support for Ubuntu 14.4.3 for Intel® Energy Profiler (SoC Watch 2.1.1).

Intel® Graphics Performance Analyzers (Intel® GPA)

• Added support for 32-bit and 64-bit applications on Android M (6.0, Marshmallow).
• Intel Graphics Performance Analyzers are now in a single package for Windows users.
• Added support for OS X 10.11 El Capitan.
• Implemented texture storage parameters modification experiment - you can now change dimensions and sample count parameters for input textures without recompiling your app.
• Can now export textures in KTX/DDS/PNG file formats.

Intel® Inspector

• Support for Fedora 23 and Ubuntu 15.10.

Update 1

FreeBSD* Target Release Notes Linux* Host Release Notes Windows* Target Release Notes

Intel® C++ Compiler

• Enhancements for offloading to Intel® Graphics Technology
• Added Intel® SIMD Data Layout Templates

Intel® System Debugger

• New options for the debugger’s “Restart” command
• NSystem Trace Viewer:
  • New "Event Distribution View" feature.
  • Several improvements in the Trace Viewer GUI.

Intel® Energy Profiler (SoC Watch)

• Added support for collection of gfx-cstate and ddr-bw metrics on platforms based on Intel® Core™ architecture.

Initial Release

FreeBSD* Target Release Notes System Requirements

Intel® C++ Compiler

• Enhanced C++11 feature support
• Enhanced C++14 feature support
• FreeBSD* support

Intel® Math Kernel Library (Intel® MKL)

• New ?GEMM_BATCH and (C/Z)GEMM3M_BATCH functions for performing multiple independent matrix-matrix multiply operation.
• New C-language version of the Intel® MKL reference manual

Intel® Integrated Performance Primitives (Intel® IPP)

• Extended optimization for Intel Atom® processors in the Computer Vision and Image Processing functions.
• Added optimization for Intel® Quark™ processors to the Cryptography function.

Intel® System Debugger

• Support for new platforms based on Airmont microarchitecture: Moorefield (Z35XX), Cherrytrail (Z8700), Braswell (N3700).
• New supported targets: 6th Generation Intel® Core™ Processor Family, Intel® 100 Series Chipset.
• For 6th Generation Intel® Core™ Processor Family :
  • Intel® Debug Extensions for WinDbg* with Intel® Processor Trace support and JTAG debug support
  • System Trace support for Intel® Trace Hub
  • Intel® Debugger for Heterogeneous Compute
  • The debugger supports 64-bit host OS systems only and requires a 64-bit Java* Runtime Environment (JRE) to operate. See System Debugger release notes for more details.

Intel® VTune™ Amplifier for Systems

• Basic Hotspots, Locks and Waits and hardware event-based stack sampling collection supported for RT kernel and RT applications for Linux* targets.
• Hardware event-based stack sampling collection supported for kernel-mode threads.
• Support for Intel Atom® x7 Z8700 & x5 Z8500/X8400 processor series (Cherry Trail) including GPU analysis.
• KVM guest OS profiling based on the Linux* Perf tool.
• Analysis of applications in a virtualization environment (KVM) for Linux* kernels (version 3.2 and higher) and QEMU (version 1.4 and higher).
• Remote event-based sampling analysis on SoFIA, leveraging an existing sampling driver on the target.

 

Component Release Notes

Release notes for individual components

• Intel® C/C++ Compiler
  Linux Host
  Windows Host
  Windows Target
• GNU* GDB
• Intel® Math Kernel Library (Intel® MKL)
• Intel® Integrated Performance Primitives (Intel® IPP)
• Intel® Threading Building Blocks (Intel® TBB)
• Intel® Data Analytics Acceleration Library (Intel® DAAL)
• OpenCL™ Tools release notes
  Developer's Guide for OpenCL™ Tools
• MRAA IO Communication Layer
  UPM Sensor and Actuator Library
• Intel® VTune™ Profiler
• Intel® SoC Watch
• Intel® Advisor
• Intel® Inspector
• Intel® System Debugger - System Debug, System Trace, and Intel® Debug Extensions for WinDBG*

Additional Documentation

To automatically receive product updates, users must register or set up their account with the Intel® Software Development Products Registration Center.

Start development or optimization with these Getting Started guides.

You can find more information about each individual component here.

For information on supported and unsupported product versions, visit Intel® System Studio Supported and Unsupported Product Versions.

For questions or technical support, visit Intel® Software Developer Support.

 

*Other names and brands may be claimed as the property of others

OpenCL and the OpenCL logo are trademarks of Apple Inc. used by permission by Khronos.