Intel® oneAPI Toolkits Installation Guide for Linux

Última atualização:10/02/2020

1. Introduction

This guide covers the installation of Intel® oneAPI Toolkits on Linux* systems. The following installation options are available:

 

Installation Option Linux
Online Installer Yes
Full package local installer Yes
Container Yes
APT Yes
YUM Yes
Zypper Yes

Important:

Important notes if you will offload to Intel GPUs:

  • If you are using Intel GPU, you will need to install the latest GPU drivers separately.
  • Check that you have fulfilled the requirements of the Intel® Graphics Compute Runtime for oneAPI Level Zero and OpenCL™ Driver. In particular, you need to have permissions to access files /dev/dri/renderD* and /dev/dri/card*. This typically means that your user account is a member of the "video" (on Ubuntu* 18, Fedora* 30, and SLES* 15 SP1) or "render" (on Ubuntu* 19 and higher, CentOS* 8, and Fedora* 31) group.  Alternatively, an administrator with sudo or root privilege can change the group owner of /dev/dri/renderD* /dev/dri/card* to a group ID used by your user base. 
  • If you plan to use the Intel® Distribution for GDB* on Linux*, make sure to configure debugger access.
  • If you have applications with long-running GPU compute workloads in native environments, you must disable the hangcheck timeout period to avoid terminating workloads.

2. Online Installer

2.1 Download the Online Installer

  1. Go to the Intel® oneAPI page.
  2. Locate the Intel® oneAPI Base Toolkit or Domain-Specific Toolkit for Specialized Workloads.

Note Some domain-specific toolkits require you to install the Intel oneAPI Base Toolkit first for full functionality.

Note To use third-party IDE, install Eclipse* on your Linux* host system before installing oneAPI Toolkits. This allows you to integrate the plugins as part of the Base Kit installation.

 

  1. Click the Get the Base Kit button or click the Domain-Specific Toolkit button for standalone capable kits.
  2. Select your OS from the drop-down list.
  3. For a web-based download, click the Configure Install button. You will also have a container and package manager options. Refer to these installation instructions.

 

  1. Click on the Download button for the Online Installer. Alternatively, you can choose the Local Installer.

Online Installer

 

NOTE: If you are targeting an FPGA and would like to install additional components for this kind of development, click on the Intel® FPGA Add-on for oneAPI Base Toolkit in the gray box. This provides an add-on, separate download. See the FPGA install instructions for detailed steps.

2.2 Run the Online Installer

  1. If the downloaded .sh file does not have "x" in file mode, use the following command to add:

    chmod +x ./l_[Toolkit Name]Kit_[version].sh
  2. Use the following command to launch the installer:

    ./l_[Toolkit Name]Kit_[version].sh
  3. Follow the installer instructions.

Once the installation is complete, verify that your toolkit has been installed to the correct installation directory:

  • root: /opt/intel/oneapi
  • user: ~/intel/oneapi

Note If you are using Intel GPU, you will need to install the latest GPU drivers separately.

3. Local Installer (Full package)

This section contains steps to install your toolkit using the Local Installer (full package).

On selected toolkit download page click Get It Now button. Select your OS and then under Local Installer on the right, click the Download button. Save the installation file to your computer.

Local Installer

You can choose to install:

3.1 Install with the GUI

  1. If the downloaded .sh file does not have "x" in file mode, use the following command to add:

    chmod +x ./l_[Toolkit Name]Kit_[version].sh
  2. Use the following command to launch the installer:
    ./l_[Toolkit Name]Kit_[version].sh
  3. Follow the installer instructions.

Once the installation is complete, verify that your toolkit has been installed to the default directory

  • root: /opt/intel/oneapi
  • user: ~/intel/oneapi

Note If you are using Intel GPU, you will need to install the latest GPU drivers separately.

3.2 Install with the Command Line

3.2.1 Silent Install

Follow the instructions below:

  1. Go to the directory where the downloaded toolkit package (.sh) is located. If the downloaded .sh file does not have "x" in file mode, use the following command to add:
    chmod +x ./l_[Toolkit Name]Kit_[version].sh
  2. Launch the installation using the following command:
    sudo ./l_[Toolkit Name]Kit_[version].sh -s -a --silent --eula accept

The installation may take some time to finish. Once the installation is complete, verify that the toolkit is installed in the default directory: /opt/intel/oneapi

3.2.2 Interactive Install

  1. Use the following command to launch the installer:
    ./l_[Toolkit Name]Kit_[version].sh -a --cli 
  2. Follow installer instructions.
  3. Once the installation is complete, verify that your toolkit has been installed to the correct directory: /opt/intel/oneapi

NOTE: If you are using Intel GPU, you will need to install the latest GPU drivers separately.

3.2.3 Custom Install oneAPI Toolkit

The following explains how to install selected oneAPI Toolkit components from the Command Line. You can specify which products or components to include when installing your toolkit from the command line. Follow the steps below:

1. Display the list of already installed products and products included in the downloaded package using the following command:

l_[Toolkit Name]Kit_[version].sh -s -a --list-products

Example of output:
ID Version Language Installed Name
==========================================================================================================
intel.oneapi.win.tbb.product 2021.1-496.beta08 false Intel(R) oneAPI Threading Building Blocks

2. Display the list of components in product of current package using the following command:

l_[Toolkit Name]Kit_[version].sh -s -a --list-components 

3. Display the list of components of any installed product on the system using the following command:

l_[Toolkit Name]Kit_[version].sh -s -a --list-components --product-id intel.oneapi.win.tbb.product --product-ver 2021.1-496.beta08

Example of output:
ID Version Language Name
==============================================================================================
intel.oneapi.win.tbb.devel 2021.1-496.beta08 Intel(R) oneAPI Threading Building Blocks

4. Run the installation with the command:

sudo ./install.sh --silent --eula accept
  • Once the installation is complete, verify that the toolkit is installed in the default directory
    • root: /opt/intel/oneapi
    • user: ~/intel/oneapi

NOTE: If you are using Intel GPU, you will need to install the latest GPU drivers separately.

4. Install Packages or Components to Different Directories

By default, when you install oneAPI Toolkit or component the installer saves a database with the location you use for the installation.  Then after this initial installation all further installations will default to this directory stored in the database.  Sometimes this is not wanted.  As an example, sometimes you may want to install oneAPI Base Toolkit in one directory and oneAPI HPC Toolkit in a completely different directory.

The installers support using a non-default installation directory on the first installation using the CUSTOM or SILENT installation methods.  However, after this initial installation, the installer leaves a database with the directory information from this initial installation. This may prevent you from installing other toolkits in directories other than the first toolkit installation.  Or perhaps you want to install components separately to different directories.

To support changing the installation directory after the initial installation there is a work-around: you can remove the installer database. Remember though that removing the database in this workaround will break the product uninstall function (as the database used by uninstall will be gone). There is a work-around for removing Toolkits or components after the database is removed: you can uninstall the product manually (shown below).

To remove the history database from the installer do this:

  • root installation:

           Go to /var/intel/installercache directory and remove the packagemanager.db database.

  • user (non-root) installation:

           Go to /home/<username>/intel/installercache directory and remove the packagemanager.db database.

Removing the database will now allow a new installation to change the installation directory used by the previous installation.  As an example, if you want to install the Base Toolkit in one directory and HPC Toolkit in another:

  • First, install oneAPI Base Toolkit in a directory of your choice: for this example we may choose /opt/base directory.
  • Next, before installing oneAPI HPC Toolkit, the history saved from the Base Toolkit is removed by deleting file /var/intel/installercache/packagemanager.db.
  • Finally, now you can install oneAPI HPC Toolkit in another directory, say for example /opt/hpc directory.

Suppose at some point in the future you want to remove the first Base Toolkit.  Using the Uninstaller you can still uninstall HPC Toolkit  but Base Toolkit won't uninstall this way since it’s history database was removed.  In this case, to uninstall Base Toolkit, remove the /opt/base directory with a command like 'rm -Rf /opt/base' to recursively remove all files under /opt/base.
 

5. Other Installation Methods

5.1 Package Managers

Follow these instructions to install from Linux* repositories:

  1. Install from yum (Internet connection required)
  2. Install from APT (Internet connection required)
  3. Install from Zypper (Internet connection required)

5.2 Docker Container

Follow the instructions to install a Docker container.

6. Set Up a System for FPGA with the Intel® PAC

There are multiple ways you can work with FPGAs:

6.1 Install the Intel® FPGA Add-on for oneAPI Base Toolkit Package

The Intel® FPGA Add-on for oneAPI Base Toolkit provides the necessary tools to compile an FPGA bitstream for deployment onto your chosen FPGA platform. It is intended for use in conjunction with the Intel® oneAPI DPC++/C++ Compiler, and cannot be used as a standalone component. Use the FPGA add-on to compile your FPGA design for deployment to FPGA hardware.

The Intel® FPGA Add-on for oneAPI Base Toolkit package includes the Intel® Quartus® Prime Pro Edition and the board support package (BSP) for its targeted FPGA platform. The Intel® Quartus® Prime Pro Edition contains the following:

  • Hybrid Placer & Global Router
  • Timing Analyzer
  • Physical Synthesis
  • Incremental Fitter Optimization
  • Interface Planner
  • Synthesis Engine
  • Platform Designer
  • Partial Reconfiguration
  • Block-Based (Hierarchical) Design

NOTE: The FPGA add-on is not required for the emulation or report generation stages of the FPGA development flow. You can complete those stages with just the Intel® oneAPI DPC++/C++ Compiler.

6.1.1 Install the Intel® FPGA Add-on for Intel® PAC with Intel® Arria® 10 GX FPGA and Intel® FPGA PAC D5005

If you have the Intel® PAC with Intel® Arria® 10 GX FPGA (a10gx) or the Intel® FPGA PAC D5005 (s10sx) on the system and you want to compile your FPGA design for deployment to FPGA hardware, perform the following steps:

  1. Download the Intel® FPGA add-on for oneAPI Base Toolkit package for a10gx and s10sx package, as described in the Online Installer section. See also Intel® FPGA Add-On for oneAPI Base Toolkit home page.
  2. Unzip the package using the following command:
    unzip intel-fpga-addon-for-oneapi-base-kit-for-a10gx-s10sx-linux.zip
  3. Run the setup.sh script in the add-on package directory, as follows:
    cd intel-fpga-addon-for-oneapi-base-kit-for-a10gx-s10sx-linux/l_intel_fpga_dpcpp_2021.1-a10gx-s10sx-beta09
    sudo ./setup.sh 

    CAUTION: To install the Intel FPGA add-on package, you need root or sudo privileges. Otherwise, the installer displays an error.

  4. Perform the following for the installer prompts:
    1. Accept the license agreement.
    2. Enter your Intel® oneAPI Base Toolkit installation path. The default path is /opt/intel/oneapi/
    3. Choose to install the prerequisite packages.
  5. Close the existing shell.
  6. Open a new shell and set up the Intel® oneAPI Base Toolkit user environment.
    cd /opt/intel/oneapi or <you own oneAPI base toolkit installation path>
    source setvars.sh
  7. [OS-specific] For systems with Red Hat Enterprise Linux (RHEL) 8 and CentOS 8, execute the following commands to install libnsl, python2, and jsonschema packages:
    yum install libnsl
    yum install python2
    pip2 install jsonschema
  8. Diagnose the oneAPI Base Toolkit installation, including the FPGA installation and environment by running the sys_check.sh script:
    cd /opt/intel/oneapi or <your own oneAPI base toolkit installation path>  
    ./sys_check.sh

    NOTE: If this diagnostic passes, the script does not print anything. If it fails, the script prints why it failed.

6.1.2 Install the Intel® FPGA Add-on for Intel® Custom Platform

Two add-on options are supported for the Custom Platform. One containing the Intel® Quartus® Prime software version 19.4 and the other containing the Intel® Quartus® Prime software version 20.2.
If you are unsure about which software your Custom Platform supports, check with your board vendor.

If you have Intel® Custom Platform on the system and you want to compile your FPGA design for deployment to FPGA hardware, perform the following steps:

  1. Download the Intel® FPGA add-on for oneAPI Base Toolkit package for Custom Platform, as described in the Online Installer section. See also Intel® FPGA Add-On for oneAPI Base Toolkit home page.
  2. Unzip the package.
    • For the Intel® Quartus® Prime software version 19.4 package, use the following command:
      unzip intel-fpga-addon-for-oneapi-base-kit-for-custom-platforms-quartus19.4-linux.zip
    • For the Intel® Quartus® Prime software version 20.2 package, use the following command:
      unzip intel-fpga-addon-for-oneapi-base-kit-for-custom-platforms-quartus20.2-linux.zip
  3. Run the setup.sh script in the add-on package directory.
    • For the Intel® Quartus® Prime software version 19.4 package, use the following command:
      cd intel-fpga-addon-for-oneapi-base-kit-for-custom-platforms-quartus19.4-linux
      cd l_intel_fpga_dpcpp_2021.1-custom-platforms-quartus19.4-beta09
      sudo ./setup.sh
    • For the Intel® Quartus® Prime software version 20.2 package, use the following command:
      cd intel-fpga-addon-for-oneapi-base-kit-for-custom-platforms-quartus20.2-linux
      cd l_intel_fpga_dpcpp_2021.1-custom-platforms-quartus20.2-beta09
      sudo ./setup.sh

      CAUTION To install the Intel FPGA add-on package, you need root or sudo privileges. Otherwise, the installer displays an error.

  4. Perform the following for the installer prompts:
    1. Accept the license agreement.
    2. Enter your Intel® oneAPI Base Toolkit installation path. By default, it is /opt/intel/oneapi/
    3. Choose to install prerequisite packages.
  5. Visit the Intel® FPGA Development Tools website and download the following two packages:

    The Intel® FPGA add-on for oneAPI Base Toolkit for Custom Platforms does not include the software stack for a10_ref and s10_ref. So, you must obtain the following software stack manually:

    • Intel FPGA OpenCL Board Support Package for Arria 10 GX Linux x86-64 TGZ
    • Intel FPGA OpenCL Board Support Package for Stratix 10 GX Linux x86-64 TGZ
  6. Unzip the packages to one of the following locations:
    • /opt/intel/oneapi/intelfpgadpcpp/2021.1-beta09/board
      or
    • <your own oneAPI base toolkit installation path>/intelfpgadpcpp/2021.1-beta09/board
  7. Close the existing shell.
  8. Open a new shell and set up the Intel® oneAPI Base Toolkit user environment.
    cd /opt/intel/oneapi or <you own oneAPI base toolkit installation path>
    source setvars.sh
  9. Diagnose the oneAPI Base Toolkit installation, including the FPGA installation and environment by running the sys_check.sh script:
    cd /opt/intel/oneapi or <your own oneAPI base toolkit installation path>  
    ./sys_check.sh

    NOTE: If this diagnostic passes, the script does not print anything. If it fails, the script prints why it failed.

6.2 Install the Intel® PAC Software Stack

This step applies only to the machine with a physical card or the runtime system. If you need an Intel® PAC with Intel Arria® 10 GX FPGA or an Intel® FPGA PAC D5005 (previously known as Intel® PAC for Intel® Stratix® 10 SX FPGA), contact the Intel sales representatives. Refer to www.intel.com/arria10pac and www.intel.com/d5005pac for more information.

Note: For systems that are used for FPGA bitstream compilation and generation, execute the steps in Install the Software Stack Shipped with Intel® FPGA add-on Package to install the correct version of the Intel PAC software stack.

Perform the following steps:

  1. Install one of the following software stacks on a system with a physical card:
    • If you are on a system with the Intel® PAC with Intel Arria® 10 GX FPGA, run the following command to install its software stack:
      aocl install intel_a10gx_pac
    • If you are on a system with the Intel® FPGA PAC D5005 (previously known as Intel® PAC for Intel® Stratix® 10 SX FPGA), run the following command to install its software stack:
      aocl install intel_s10sx_pac

CAUTION: Avoid running the aocl install command with secure boot enabled. Otherwise, an error is displayed since it is currently not supported.

CAUTION: The software stack for Intel® PAC with Intel Arria® 10 GX FPGA and Intel® FPGA PAC D5005 (previously known as Intel® PAC for Intel® Stratix® 10 SX FPGA) cannot be simultaneously installed on the same system. If you have installed one of them and want to install the other one, you must uninstall the former one first by either running the aocl uninstall intel_a10gx_pac or aocl uninstall intel_s10sx_pac command.

  1. Accept the installation prompts to proceed with installing the Intel® PAC software stack.

NOTE: You must ensure that the FPGA development board is plugged into the system. Verify it using the aocl diagnose command. This diagnostic should be done only once per installation.

TIP: The installer attempts to install the prerequisite packages. If it fails to install a package (for example, due to version conflict or repository incomplete), you can either manually install that package or run the sudo apt -fix-broken install command.

6.2.1 (Caveat) Install the Correct Version of Intel® PAC Software Stack

Both Intel® oneAPI Base Toolkit and Intel® FPGA add-on contains Intel® PAC software stack. The software stack shipped with Intel® oneAPI Base Toolkit has limited functionalities, while the software stack shipped with Intel® FPGA add-on has full functionalities:

Supported Functionality Intel® PAC Software Stack in Intel® oneAPI Base Toolkit Intel® PAC Software Stack in Intel® FPGA Add-on
Emulation
Report generation
FPGA runtime
FPGA bitstream generation  

Running steps to install the software stack from section 5.2 before installing the FPGA Add-on in section 5.1 causes the PAC software stack in Base Toolkit to be used by default.
If you want to compile your FPGA design for deployment to FPGA hardware, you must make sure the Intel® PAC software stack shipped with Intel® FPGA add-on has been installed. You can check the installed software stack by running the aoc -list-board-packages command. The installed software stacks are listed in the Installed board packages section in the command output.

To uninstall a previously installed software stack, refer to the uninstall instructions.

6.2.2 Install the Software Stack Shipped with Intel® FPGA Add-on for oneAPI Base Toolkit Package

To install the software stack shipped with the Intel® FPGA Add-on for oneAPI Base Toolkit package, perform the following steps:

  1. Follow the steps in Install the Intel® FPGA Add-on for oneAPI Base Toolkit Package to install the FPGA add-on package.
  2. Setup the Intel® oneAPI base toolkit user environment by running the following commands:
    cd /opt/intel/oneapi or <you own oneAPI base toolkit installation path>
    source setvars.sh
  3. Install the Intel® PAC software stack.

6.2.3 Update the Firmware on your Intel® PAC with Intel® Arria® 10 GX FPGA

If you used your Intel® PAC with Intel® Arria® 10 GX FPGA with the 2021.1-beta06 or a lower version of the Intel® FPGA Add-on in the past, you must update the firmware in order for the FPGA to work with the 2021.1-beta07 and later versions of the Intel® PAC software stack.

Use the instructions in Identifying the Flash Image and BMC Firmware to identify your firmware version. If you need to update the firmware, use the instructions in Updating the FIM and BMC Firmware.

7. Install Intel GPU Drivers

If you are using Intel GPU, you will need to install the latest GPU drivers (Intel® oneAPI Level Zero and OpenCL™ Driver) separately. Follow the instructions on this new site General Purpose GPU Drivers for Linux* Operating Systems to install the latest Intel GPU drivers.

  • Check that you have fulfilled the requirements of the Intel® Graphics Compute Runtime for oneAPI Level Zero and OpenCL™ Driver. In particular, you need to have permissions to access files /dev/dri/renderD* and /dev/dri/card*. This typically means that your user account is a member of the "video" (on Ubuntu* 18, Fedora* 30, and SLES* 15 SP1) or "render" (on Ubuntu* 19 and higher, CentOS* 8, and Fedora* 31) group.  Alternatively, an administrator with sudo or root privilege can change the group owner of /dev/dri/renderD* /dev/dri/card* to a group ID used by your user base. 
  • For GPU compute workloads, non-root users do not have access to the GPU device. If you do not add a non-root user to the video group, all binaries compiled for a GPU will fail during execution. To fix this problem, add the non-root user to the video group:
sudo usermod -a -G video <username>
  • If you plan to use the Intel® Distribution for GDB* on Linux*, make sure to configure debugger access.
  • If you have applications with long-running GPU compute workloads in native environments, you must disable the hangcheck timeout period to avoid terminating workloads.

8. GPU: Disable Hangcheck

This section applies only to applications with long-running GPU compute workloads in native environments. It is not recommended for virtualizations or other standard usages of GPU, such as gaming.

A workload that takes more than four seconds for GPU hardware to execute is a long-running workload. By default, individual threads that qualify as long-running workloads are considered hung and are terminated. By disabling the hangcheck timeout period, you can avoid this problem.

NOTE: If the system is rebooted, hangcheck is automatically enabled. You must disable hangcheck again after every reboot or follow the directions below to disable hangcheck persistently (across multiple reboots). Please re-run this GPU Hangcheck disable with reboot fix if you update (or auto-update) the kernel.

To disable hangcheck until the next reboot:

sudo sh -c "echo N> /sys/module/i915/parameters/enable_hangcheck"

To disable hangcheck across multiple reboots:

NOTE: If the kernel is updated, hangcheck is automatically enabled. Run the procedure below after every kernel update to ensure hangcheck is disabled.

  1. Open a terminal.
  2. Open the grub file in /etc/default.
  3. In the grub file, find the line GRUB_CMDLINE_LINUX_DEFAULT="".
  4. Enter this text between the quotes (""):
    i915.enable_hangcheck=0
  5. Run this command.
    sudo update-grub
  6. Reboot the system. Hangcheck remains disabled.

9. Install OpenCL™ Offline Compiler (OCLOC)

This topic targets Linux* users who want to use Ahead-of-Time (AOT) Compilation to generate binaries for one or multiple selected devices. For details about the AOT feature, please reference the Intel® oneAPI DPC++/C++ Compiler Developer Guide and Reference.

There are currently three operating systems supported, RHEL* 8.0, Ubuntu* 18.04, and Ubuntu 19.10. The directions to install the OCLOC are listed below.

9.1 RHEL 8.0

To access OCLOC packages, add the repository to your system and then install the packages using dnf (Important: root permission is required)

cat << EOF | sudo tee /etc/yum.repos.d/intel-graphics.repo
[intel-graphics]
name=Intel Graphics Drivers Repository baseurl=https://repositories.intel.com/graphics/rhel/8.0/
enabled=1
gpgcheck=0
EOF
sudo dnf install intel-ocloc

9.2 Ubuntu 18.04

Note For Ubuntu 18.04, the distribution is bionic in the repository line below.

To access OCLOC packages, add the repository to your system and then install the packages using apt (Note: root permission is required)

cat << EOF | sudo tee /etc/apt/sources.list.d/intel-graphics.list
deb [trusted=yes arch=amd64] https://repositories.intel.com/graphics/ubuntu bionic main
EOF
sudo apt-get update
sudo apt-get install intel-ocloc

9.3 Ubuntu 19.10

Note For Ubuntu 19.10, the distribution is eoan in the repository line below.

To access OCLOC packages, add the repository to your system and then install the packages using apt (Note: root permission is required)

cat << EOF | sudo tee /etc/apt/sources.list.d/intel-graphics.list
deb [trusted=yes arch=amd64] https://repositories.intel.com/graphics/ubuntu eoan main
EOF
sudo apt-get update
sudo apt-get install intel-opencl-icd

10. Get Started

For instructions on how to build and run Samples see Get Started With the Intel oneAPI Toolkits for Linux.

10.1 Code Samples

Intel oneAPI code samples are available for download on GitHub* to help you get started.

11. Uninstall oneAPI Toolkits and components

11.1 Uninstall Intel PAC Card

To uninstall an installed software stack, you can run one of the following commands:

  • Intel® PAC for Intel® Arria® 10 GX FPGA
    aocl uninstall intel_a10gx_pac
  • Intel® FPGA PAC D5005
    aocl uninstall intel_s10sx_pac 

11.2 Uninstall oneAPI Toolkits

11.2.1 Uninstall Using GUI

Use the following commands to uninstall oneAPI Toolkits:

cd /opt/intel/oneapi/installer 
sudo ./installer
  • Select "Remove" to remove the toolkit
  • Use Installer dashboard dialog with the list of already installed products (toolkits) to Modify, Repair or Remove each toolkit separately.

11.2.2 Use Silent CLI

Use the following commands to uninstall oneAPI Toolkits:

  1. Display the list of already installed products and products included in the downloaded package using the following command:
l_[Toolkit Name]Kit_[version].sh -s -a --list-products

 Example of output:
ID Version Language Installed Name
==========================================================================================================
intel.oneapi.win.tbb.product 2021.1-496.beta08 false Intel(R) oneAPI Threading Building Blocks

        2. Uninstall selected product:

cd /opt/intel/oneapi/installer
sudo ./installer --action remove --product-id intel.oneapi.win.tbb.product --product-ver 2021.1-496.beta08

11.2.3 Use Linux Package Manager

  • APT
sudo apt autoremove <package_name>
  • YUM
sudo yum autoremove <package_name>

11.3 Uninstall PyTorch* and TensorFlow* (Part of Intel® AI Analytics Toolkit)

11.3.1 Uninstall PyTorch

  1. Deactivate the pytorch environment, if activated.
  2. Uninstall using the following commands:
    conda remove -p $<install_dir>/pytorch/1.1.0/ --all
    rm -rf $<install_dir>/pytorch

11.3.2 Uninstall TensorFlow

  1. Deactivate the tensorflow environment, if activated.
  2. Uninstall using the following commands:
conda remove -p $<install_dir>/tensorflow/1.14.0/ --all
rm -rf $<install_dir>/tensorflow

11.4 Uninstall the Intel® FPGA Add-on for oneAPI Base Toolkit

To uninstall a specific version of the Intel® FPGA Add-on for oneAPI Base Toolkit, use the following commands:

sudo chmod +w $<install_dir>/intelfpgadpcpp/<version>/ -R
sudo rm -rf $<install_dir>/intelfpgadpcpp/<version>/

 


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.

Informações de produto e desempenho

1

Os compiladores da Intel podem ou não otimizar para o mesmo nível de microprocessadores não Intel no caso de otimizações que não são exclusivas para microprocessadores Intel. Essas otimizações incluem os conjuntos de instruções SSE2, SSE3 e SSSE3, e outras otimizações. A Intel não garante a disponibilidade, a funcionalidade ou eficácia de qualquer otimização sobre microprocessadores não fabricados pela Intel. As otimizações que dependem de microprocessadores neste produto são destinadas ao uso com microprocessadores Intel. Algumas otimizações não específicas da microarquitetura Intel são reservadas para os microprocessadores Intel. Consulte os Guias de Usuário e Referência do produto aplicáveis para obter mais informações sobre os conjuntos de instruções específicos cobertos por este aviso.

Revisão do aviso #20110804