Skip to content

Commit

Permalink
doc: added separate README for binary packages
Browse files Browse the repository at this point in the history
  • Loading branch information
vpirogov committed Dec 2, 2019
1 parent 79f204c commit 28f4c96
Show file tree
Hide file tree
Showing 3 changed files with 194 additions and 83 deletions.
8 changes: 7 additions & 1 deletion CMakeLists.txt
Original file line number Diff line number Diff line change
Expand Up @@ -105,6 +105,11 @@ configure_file(
)
include_directories(${PROJECT_BINARY_DIR}/include)

configure_file(
"${PROJECT_SOURCE_DIR}/README.binary.in"
"${PROJECT_BINARY_DIR}/README"
)

if(DNNL_INSTALL_MODE STREQUAL "BUNDLE" AND NOT DEFINED CMAKE_INSTALL_LIBDIR)
# define CMAKE_INSTALL_LIBDIR as "lib" in the case of bundle
set(CMAKE_INSTALL_LIBDIR "lib")
Expand All @@ -119,8 +124,9 @@ add_subdirectory(tests)

if(DNNL_INSTALL_MODE STREQUAL "BUNDLE")
install(FILES LICENSE DESTINATION ${CMAKE_INSTALL_PREFIX})
install(FILES README.md DESTINATION ${CMAKE_INSTALL_PREFIX})
install(FILES ${PROJECT_BINARY_DIR}/README DESTINATION ${CMAKE_INSTALL_PREFIX})
else()
# Cannot use CMAKE_INSTALL_DOCDIR since it uses PROJECT_NAME and not LIB_NAME
install(FILES LICENSE DESTINATION ${CMAKE_INSTALL_DATAROOTDIR}/doc/${LIB_NAME})
install(FILES ${PROJECT_BINARY_DIR}/README DESTINATION ${CMAKE_INSTALL_DATAROOTDIR}/doc/${LIB_NAME})
endif()
115 changes: 115 additions & 0 deletions README.binary.in
Original file line number Diff line number Diff line change
@@ -0,0 +1,115 @@
Deep Neural Network Library (DNNL)
==================================

Deep Neural Network Library (DNNL) is an
open-source performance library for deep learning applications. The library
includes basic building blocks for neural networks optimized
for Intel Architecture Processors and Intel Processor Graphics.

This package contains DNNL v@PROJECT_VERSION@ (@DNNL_VERSION_HASH@).

You can find information about the latest version and release notes
at DNNL Github (https://github.com/intel/mkl-dnn/releases).

Documentation
-------------

* Developer guide (https://intel.github.io/mkl-dnn/v@DNNL_VERSION_MAJOR@.@DNNL_VERSION_MINOR@)
explains programming model, supported functionality, details of primitives
implementations and includes annotated examples.
* API reference (https://intel.github.io/mkl-dnn/v@DNNL_VERSION_MAJOR@.@DNNL_VERSION_MINOR@/modules.html)
provides comprehensive reference of the library API.

System Requirements
-------------------

DNNL supports systems based on Intel 64 architecture or
compatible processors.

The library is optimized for the following CPUs:
* Intel Atom processor with Intel SSE4.1 support
* 4th, 5th, 6th, 7th, and 8th generation Intel Core(TM) processor
* Intel Xeon(R) processor E3, E5, and E7 family (formerly Sandy Bridge,
Ivy Bridge, Haswell, and Broadwell)
* Intel Xeon Phi(TM) processor (formerly Knights Landing and Knights Mill)
* Intel Xeon Scalable processor (formerly Skylake and Cascade Lake)
* future Intel Xeon Scalable processor (code name Cooper Lake)

DNNL detects instruction set architecture (ISA) in the runtime and uses
just-in-time (JIT) code generation to deploy the code optimized
for the latest supported ISA.

The library is optimized for the following GPUs:
* Intel HD Graphics
* Intel UHD Graphics
* Intel Iris Plus Graphics

## Linux

Common dependencies:
* glibc 2.12 or later
* GCC 4.8 or later

Runtime specific dependencies:

| Runtime configuration | Requirements |
| --------------------- | ---------------------------------------------------- |
| `cpu_gomp` | No additional requirements |
| `cpu_iomp` | Intel OpenMP runtime |
| `cpu_tbb` | Threading Building Blocks 2017 or later |

## Windows

Common dependencies:
* Microsoft Visual C++ Redistributable 2015 or later

Runtime specific dependencies:

| Runtime configuration | Requirements |
| --------------------- | -----------------------------------------------------|
| `cpu_vcomp` | No additional requirements |
| `cpu_iomp` | Intel OpenMP runtime |
| `cpu_tbb` | Threading Building Blocks 2017 or later |

## macOS

Common dependencies:
* macOS 10.13 (High Sierra) or later

Runtime specific dependencies:

| Runtime configuration | Requirements |
| --------------------- | -----------------------------------------------------|
| `cpu_iomp` | Intel OpenMP runtime |
| `cpu_tbb` | Threading Building Blocks 2017 or later |

Support
-------

Please submit your questions, feature requests, and bug reports on the
GitHub issues page (https://github.com/intel/mkl-dnn/issues/new/choose).

You may reach out to project maintainers privately at [email protected].

WARNING:
The following functionality has preview status and might be changed without
prior notification in future releases:
* Primitive cache

License
-------

DNNL is licensed under Apache License Version 2.0. This
software includes components with separate copyright notices and license
terms. Your use of the source code for these components is subject to the terms
and conditions of the following licenses.

3-clause BSD license:
* Xbyak (https://github.com/herumi/xbyak)
* ittnotify (https://github.com/intel/IntelSEAPI)
* CMake (https://github.com/Kitware/CMake)

Boost Software License, Version 1.0:
* Boost C++ Libraries (https://www.boost.org/)

See accompanying LICENSE file for full license text and copyright notices.
154 changes: 72 additions & 82 deletions README.md
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
Deep Neural Network Library (DNNL)
========================================================================
==================================

> **Note**
>
Expand Down Expand Up @@ -35,6 +35,14 @@ applications enabled with DNNL:
* [PyTorch\*](https://pytorch.org/)
* [Tensorflow\*](https://www.tensorflow.org)

# Documentation

* [Developer guide](https://intel.github.io/mkl-dnn) explains programming
model, supported functionality, details of primitives implementations and
includes annotated examples.
* [API reference](https://intel.github.io/mkl-dnn/modules.html) provides
comprehensive reference of the library API.

# Installation

Pre-built binaries for Linux\*, Windows\*, and macOS\* are available for download
Expand Down Expand Up @@ -67,56 +75,8 @@ details on CPU and GPU runtimes.
If the configuration you need is not available, you can
[build the library from source](http://intel.github.io/mkl-dnn/dev_guide_build.html).

# Contributing
We welcome community contributions to DNNL. If you have an idea on how
to improve the library:

* For changes impacting the public API, submit
an [RFC pull request](CONTRIBUTING.md#RFC_pull_requests).
* Ensure that the changes are consistent with the
[code contribution guidelines](CONTRIBUTING.md#code_contribution_guidelines)
and [coding style](CONTRIBUTING.md#coding_style).
* Ensure that you can build the product and run all the examples with your
patch.
* Submit a [pull request](https://github.com/intel/mkl-dnn/pulls).

For additional details, see [contribution guidelines](CONTRIBUTING.md).

# Support

> **WARNING**
>
> The following functionality has preview status and might be changed without prior notification in future releases:
> * [Primitive cache](https://intel.github.io/mkl-dnn/dev_guide_primitive_cache.html)
Please submit your questions, feature requests, and bug reports on the
[GitHub issues](https://github.com/intel/mkl-dnn/issues) page.

# License
DNNL is licensed under
[Apache License Version 2.0](http://www.apache.org/licenses/LICENSE-2.0). This
software includes the following third-party components:
* [Xbyak](https://github.com/herumi/xbyak) distributed under
[3-clause BSD licence](src/cpu/xbyak/COPYRIGHT)
* [gtest](https://github.com/google/googletest) distributed under
[3-clause BSD license](tests/gtests/gtest/LICENSE)
* [ittnotify](https://github.com/intel/IntelSEAPI) distributed under
[3-clause BSD license](src/cpu/jit_utils/jitprofiling/LICENSE.BSD)
* [Cmake](https://github.com/Kitware/CMake) distributed under
[3-clause BSD license](cmake/Copyright.txt)
* [MathJax](https://github.com/mathjax/MathJax) distributed under
[Apache License Version 2.0](doc/assets/mathjax/LICENSE)
* [Boost C++ Libraries](https://www.boost.org/) distributed under
[Boost Software License, Version 1.0](src/common/LICENSE_1_0)

# Documentation
* [Developer guide](https://intel.github.io/mkl-dnn) explains programming
model, supported functionality, details of primitives implementations and
includes annotated examples.
* [API reference](https://intel.github.io/mkl-dnn/modules.html) provides
comprehensive reference of the library API.

# System Requirements

DNNL supports systems based on Intel 64 architecture or
compatible processors.

Expand All @@ -139,6 +99,7 @@ The library is optimized for the following GPUs:
* Intel Iris Plus Graphics

## Requirements for Building from Source

DNNL supports systems meeting the following requirements:
* Operating sytem with Intel 64 architecture support
* C++ compiler with C++11 standard support
Expand All @@ -150,6 +111,7 @@ Configurations of CPU and GPU engines may introduce additional build time
dependencies.

### CPU Engine

Intel Architecture Processors and compatible devices are supported by the
DNNL CPU engine. The CPU engine is built by default and cannot
be disabled at build time. The engine can be configured to use the OpenMP or
Expand All @@ -163,6 +125,7 @@ Some implementations rely on OpenMP 4.0 SIMD extensions, and we recommend using
the Intel C++ Compiler for the best performance results.

### GPU Engine

Intel Processor Graphics is supported by the DNNL GPU engine. The GPU
engine is disabled in the default build configuration. The following
additional requirements apply when GPU engine is enabled:
Expand All @@ -171,10 +134,12 @@ additional requirements apply when GPU engine is enabled:
with Intel subgroups extension support

### Runtime Dependencies

When DNNL is built from source, the library runtime dependencies
and specific versions are defined by the build environment.

#### Linux

Common dependencies:
* System C/C++ runtime (libc.so, libstdc++.so)
* Dynamic Linking Library (libdl.so)
Expand All @@ -183,39 +148,42 @@ Common dependencies:

Runtime specific dependencies:

| Runtime configuration | Compiler | Dependency
| :----------------------- | :---------------------------- | :---------
| Runtime configuration | Compiler | Dependency
| :--------------------- | :---------------------------- | :---------
| `DNNL_CPU_RUNTIME=OMP` | GCC | GNU OpenMP runtime (libgomp.so)
| `DNNL_CPU_RUNTIME=OMP` | Intel C/C++ Compiler | Intel OpenMP runtime (libiomp5.so)
| `DNNL_CPU_RUNTIME=OMP` | Clang | Intel OpenMP runtime (libiomp5.so)
| `DNNL_CPU_RUNTIME=TBB` | any | Threading Building Blocks (libtbb.so)
| `DNNL_GPU_RUNTIME=OCL` | any | Intel Graphics Compute Runtime for OpenCL (libOpenCL.so)

#### Windows

Common dependencies:
* Microsoft Visual C++ Redistributable (msvcrt.dll)

Runtime specific dependencies:

| Runtime configuration | Compiler | Dependency
| :----------------------- | :---------------------------- | :---------
| Runtime configuration | Compiler | Dependency
| :--------------------- | :---------------------------- | :---------
| `DNNL_CPU_RUNTIME=OMP` | Microsoft Visual C++ Compiler | No additional requirements
| `DNNL_CPU_RUNTIME=OMP` | Intel C/C++ Compiler | Intel OpenMP runtime (iomp5.dll)
| `DNNL_CPU_RUNTIME=TBB` | any | Threading Building Blocks (tbb.dll)
| `DNNL_GPU_RUNTIME=OCL` | any | Intel Graphics Driver (OpenCL.dll)

#### macOS

Common dependencies:
* System C/C++ runtime (libc++.dylib, libSystem.dylib)

Runtime specific dependencies:

| Runtime configuration | Compiler | Dependency
| :----------------------- | :---------------------------- | :---------
| Runtime configuration | Compiler | Dependency
| :--------------------- | :---------------------------- | :---------
| `DNNL_CPU_RUNTIME=OMP` | Intel C/C++ Compiler | Intel OpenMP runtime (libiomp5.dylib)
| `DNNL_CPU_RUNTIME=TBB` | any | Threading Building Blocks (libtbb.dylib)

### Validated Configurations

CPU engine was validated on RedHat\* Enterprise Linux 7 with
* GNU Compiler Collection 4.8, 5.4, 6.1, 7.2, and 8.1
* Clang\* 3.8.0
Expand Down Expand Up @@ -249,40 +217,62 @@ on Windows Server 2019 with

## Requirements for Pre-built Binaries

### Linux
Common dependencies:
* GCC 4.8 or later
See README included into corresponding binary package.

Runtime specific dependencies:
# Support

| Runtime configuration | Requirements
| :-------------------- | :-----------
| `cpu_gomp` | No additional requirements
| `cpu_iomp` | Intel OpenMP runtime for Intel C/C++ Compiler 17.0 or later
| `cpu_tbb` | Threading Building Blocks 2017 or later
Please submit your questions, feature requests, and bug reports on the
[GitHub issues](https://github.com/intel/mkl-dnn/issues/new/choose) page.

### Windows
Common dependencies:
* Microsoft Visual C++ Redistributable 2015 or later
You may reach out to project maintainers privately at [email protected].

Runtime specific dependencies:
> **WARNING**
>
> The following functionality has preview status and might be changed without
> prior notification in future releases:
> * [Primitive cache](https://intel.github.io/mkl-dnn/dev_guide_primitive_cache.html)
| Runtime configuration | Requirements
| :-------------------- | :-----------
| `cpu_vcomp` | No additional requirements
| `cpu_iomp` | Intel OpenMP runtime for Intel C/C++ Compiler 17.0 or later
| `cpu_tbb` | Threading Building Blocks 2017 or later
# Contributing

### macOS
Common dependencies:
* macOS 10.13 (High Sierra) or later
We welcome community contributions to DNNL. If you have an idea on how
to improve the library:

Runtime specific dependencies:
* For changes impacting the public API, submit
an [RFC pull request](CONTRIBUTING.md#RFC_pull_requests).
* Ensure that the changes are consistent with the
[code contribution guidelines](CONTRIBUTING.md#code_contribution_guidelines)
and [coding style](CONTRIBUTING.md#coding_style).
* Ensure that you can build the product and run all the examples with your
patch.
* Submit a [pull request](https://github.com/intel/mkl-dnn/pulls).

For additional details, see [contribution guidelines](CONTRIBUTING.md).

This project is intended to be a safe, welcoming space for collaboration, and
contributors are expected to adhere to the
[Contributor Covenant](CODE_OF_CONDUCT.md) code of conduct.

# License

DNNL is licensed under
[Apache License Version 2.0](http://www.apache.org/licenses/LICENSE-2.0). This
software includes components with separate copyright notices and license
terms. Your use of the source code for these components is subject to the terms
and conditions of the following licenses.

3-clause BSD license:
* [Xbyak](https://github.com/herumi/xbyak)
* [gtest](https://github.com/google/googletest)
* [ittnotify](https://github.com/intel/IntelSEAPI)
* [CMake](https://github.com/Kitware/CMake)

Apache License Version 2.0:
* [MathJax](https://github.com/mathjax/MathJax)

Boost Software License, Version 1.0:
* [Boost C++ Libraries](https://www.boost.org/)

| Runtime configuration | Requirements
| :-------------------- | :-----------
| `cpu_iomp` | Intel OpenMP runtime for Intel C/C++ Compiler 17.0 or later
| `cpu_tbb` | Threading Building Blocks 2017 or later
See accompanying [LICENSE](LICENSE) file for full license text and copyright notices.

--------

Expand Down

0 comments on commit 28f4c96

Please sign in to comment.