Installing Intel® Performance Libraries and Intel® Distribution for Python* Using YUM Repository

This page provides general installation and support notes about the Community forum supported Intel® Performance Libraries and Intel® Distribution for Python* as they are distributed via the YUM repositories described below.

These software development tools are also available as part of the Intel® Parallel Studio XE and Intel® System Studio products. These products include enterprise-level Intel® Online Service Center support.

Setting up the Repository

Here is how to install the Intel YUM Repository. [Note: You must be logged in as root to set up and install the repository]

  1. Add the repositories in two ways:
    • Add all Intel® Performance Libraries and Intel® Distribution for Python* repositories at once:
      sudo yum-config-manager --add-repo https://yum.repos.intel.com/setup/intelproducts.repo
      You can enable or disable repositories in the intelproducts.repo file by setting the value of the enabled directive to 1 or 0 as required.
    • Add an individual product:
      • Intel® Math Kernel Library (Intel® MKL):
        sudo yum-config-manager --add-repo https://yum.repos.intel.com/mkl/setup/intel-mkl.repo
      • Intel® Integrated Performance Primitives (Intel® IPP):
        sudo yum-config-manager --add-repo https://yum.repos.intel.com/ipp/setup/intel-ipp.repo
      • Intel® Threading Building Blocks (Intel® TBB):
        sudo yum-config-manager --add-repo https://yum.repos.intel.com/tbb/setup/intel-tbb.repo
      • Intel® Data Analytics Acceleration Library (Intel® DAAL):
        sudo yum-config-manager --add-repo https://yum.repos.intel.com/daal/setup/intel-daal.repo
      • Intel® MPI Library:
        sudo yum-config-manager --add-repo https://yum.repos.intel.com/mpi/setup/intel-mpi.repo
      • Intel® Distribution for Python*:
        sudo yum-config-manager --add-repo https://yum.repos.intel.com/intelpython/setup/intelpython.repo
  2. Import the gpg public key for the repository
    sudo rpm --import https://yum.repos.intel.com/intel-gpg-keys/GPG-PUB-KEY-INTEL-SW-PRODUCTS-2019.PUB

Intel® Performance Libraries and Intel® Distribution for Python* versions available in the repository

<COMPONENT>

<VERSION>

<UPDATE>

<BUILD_NUM>

Intel® MKL

2017
2017
2017
2018
2018
2018
​2018
​2018
​2019
2019
2019
2019
2019

2
3
4
0
​1
2
​3
4
0
1
2
3
4

050
056
061
033
038
046
​051
057
​045
053
057
062
070

Intel® IPP

2017
2017
2018
2018
2018
​2018
2018
2019
2019
2019
2019
2019

2
3
0
1
2
3
4
0
1
2
3
4

050
056
033
038
046
051
​057
045
053
057
062
070

Intel® TBB

2017
2017
2018
2018
2018
​2018
2019
2019
2019
2019

6
8
0
2
​4
6
0
2
4
6

056
061
033
046
051
​057
045
057
062
070

Intel® DAAL

2017
2017
2018
2018
2018
2018
2019
2019
2019
2019
2019

3
4
0
1
2
3
0
1
2
3
4

056
061
033
038
046
​051
045
053
057
062
070

Intel® MPI Library

2018
​2018
2018
2019
2019
2019
2019
2019

2
3
4
0
1
2
3
4

046
051
057
045
053
057
062
070

Intel® Distribution for Python*

2017
2017

2018
2018
2018
2018
2019
2019
2019
2019
2019

2
3

0
1
2
3
0
1
2
3
4

045
053 (python2)
052 (python3)
018
023
037
​039
047
058
066
075
088

By downloading Intel® Performance Libraries and Intel® Distribution for Python* you agree to the terms and conditions stated in the End-User License Agreement (EULA).

Installing the library and Python packages using the YUM Package Manager

The following variables are used in the installation commands:
<PYTHON_VERSION>: 2, 3
<VERSION>: 2018, ...
<UPDATE>: 0, 1, 2, ...
<BUILD_NUMBER>: build number, check the list in previous section
<COMPONENT>: a component name from the list of available components below

Component

<COMPONENT>

Intel® Math Kernel Library

intel-mkl
intel-mkl-32bit
intel-mkl-64bit

Intel® Integrated Performance Primitives

intel-ipp
intel-ipp-32bit
intel-ipp-64bit

Intel® Threading Building Blocks

intel-tbb

Intel® Data Analytics Acceleration Library

intel-daal
intel-daal-32bit
intel-daal-64bit

Intel® MPI Library

intel-mpi

Intel® Distribution for Python*

intelpython2
intelpython3

How do I install a particular version?

  1. To install a particular version of one of the Intel® Performance Libraries:
    yum install <COMPONENT>-<VERSION>.<UPDATE>-<BUILD_NUMBER>

    Example:

    yum install intel-mkl-2018.2-046
  2. To install a particular language version of the Intel® Distribution for Python*:
    yum install intelpython<PYTHON_VERSION>

    Example:

    yum install intelpython3
  3. To specify which version of the Intel® Distribution for Python* to install:
    yum install intelpython<PYTHON_VERSION>-<VERSION>.<UPDATE>-<BUILD_NUMBER>

    Example:

    yum install intelpython3-2018.2-037

By downloading Intel® Performance Libraries and Intel® Distribution for Python* you agree to the terms and conditions stated in the End-User License Agreement (EULA).

How do I uninstall a particular version?

  1. To uninstall one of the Intel® Performance Libraries:
    yum autoremove <COMPONENT>-<VERSION>.<UPDATE>-<BUILD_NUMBER>

    Example:

    yum autoremove intel-mkl-2018.2-046
  2. To uninstall the Intel® Distribution for Python*:
    yum remove intelpython<PYTHON_VERSION>

    Example:

    yum remove intelpython3

Troubleshooting

Failures when trying to update Intel YUM repository

In certain cases, an update through the Intel YUM repositories may fail due to a caching issue. Please follow these steps in order to disable caching in YUM and clean up your local repo prior to update.

  1. Disable caching in YUM config file '/etc/yum.conf':
    http_caching=none
  2. Delete any temporary YUM files:
    rm -r /var/tmp/yum*
  3. Cleanup your local YUM repository:
    yum clean metadata
    yum clean all
  4. Reboot your machine
  5. Now re-attempt the YUM update:
    yum update

Failures when installing specific Intel package version on Red Hat*/CentOS* 7 systems

Due to a known bug in YUM 3.x (which is the default package manager in Red Hat*/CentOS* 7), you may see failures when trying to install specific versions of the Intel Performance Libraries or Intel Python packages. The bug is fixed in YUM 4 so we recommend you update your local instance of the client to the latest YUM version.

If you're running YUM 3.x or older, you will only be able to install the latest version available of the Intel products.

Have Questions?

Check out the FAQ
See our Get Help page for your support options.

Для получения подробной информации о возможностях оптимизации компилятора обратитесь к нашему Уведомлению об оптимизации.