YASK--Yet Another Stencil Kernel
- New YASK users may want to start with the YASK tutorial.
- Users with existing YASK-based code may want to jump to the backward-compatibility notices.
YASK is a framework to rapidly create high-performance stencil code including optimizations and features such as
- Vector-folding to increase data reuse via non-traditional data layout,
- Multi-level OpenMP parallelism to exploit multiple cores and threads,
- Scaling to multiple sockets and nodes via MPI with overlapped communication and compute, and
- Spatial tiling with automatically-tuned block sizes,
- Temporal tiling in multiple dimensions to further increase cache locality,
- APIs for C++ and Python: API documentation.
YASK contains a domain-specific compiler to convert scalar stencil code to SIMD-optimized code for Intel(R) Xeon Phi(TM) and Intel(R) Xeon(R) processors.
Supported Platforms and Processors:
- 64-bit Linux.
- Intel(R) Xeon(R) processors supporting the AVX, AVX2, or CORE_AVX512 instruction sets.
- Intel(R) Xeon Phi(TM) x200-family processors supporting the MIC_AVX512 instruction set.
- Intel(R) Xeon Phi(TM) x100-family coprocessors supporting the Knights-Corner instruction set (no longer tested).
- Intel(R) Parallel Studio XE Cluster Edition for Linux
for multi-socket and multi-node operation or
Intel(R) Parallel Studio XE Composer Edition for C++ Linux
for single-socket only
(2018 or later; 2019.3 or later recommended).
- There was an issue in Intel(R) MPI versions 2019.1 and 2019.2 that
caused the application to crash when allocating very
large shared-memory (shm) regions, so those
versions are not recommended when using the
-use_shmfeature. This issue was resolved in MPI version 2019.3.
- If you are using g++ version 8.x or later, Intel(R) C++ version 2019.x or later is required.
- Building a YASK kernel with the Gnu C++ compiler is possible. Limited testing with g++ 8.2.0 shows the "iso3dfd" kernel runs about 30% slower compared to the same kernel built with the Intel C++ compiler. Older Gnu C++ compilers can produce kernels that run many times slower.
- There was an issue in Intel(R) MPI versions 2019.1 and 2019.2 that caused the application to crash when allocating very large shared-memory (shm) regions, so those versions are not recommended when using the
- Gnu C++ compiler, g++ (4.9.0 or later; 8.2.0 or later recommended).
- Linux libraries
- Perl (5.010 or later).
- Gnu make.
- Bash shell.
- Optional utilities and their purposes:
gindentutility, used automatically during the build process to make the generated code easier for humans to read. You'll get a warning when running
makeif one of these doesn't exist. Everything will still work, but the generated code will be difficult to read. Reading the generated code is only necessary for debug or curiosity.
- SWIG (3.0.12 or later), http://www.swig.org, for creating the Python interface.
- Python 2 (2.7.5 or later) or 3 (3.6.1 or later), https://www.python.org/downloads, for creating and using the Python interface.
- Doxygen (1.8.11 or later), http://doxygen.org, for creating updated API documentation. If you're not changing the API documentation, you can view the existing documentation at the link at the top of this page.
- Graphviz (2.30.1 or later), http://www.graphviz.org, for rendering stencil diagrams.
- Intel(R) Software Development Emulator, https://software.intel.com/en-us/articles/intel-software-development-emulator, for functional testing if you don't have native support for any given instruction set.
Backward-compatibility notices, including changes in default behavior:
- Version 2.20.00 added checking of the step-dimension index value in the
yk_grid::get_element()and similar APIs. Previously, invalid values silently "wrapped" around to valid values. Now, by default, the step index must be valid when reading, and the valid step indices are updated when writing. The old behavior of silent index wrapping may be restored via
set_step_wrap(true). The default for all
strict_indicesAPI parameters is now
trueto catch more programming errors and increase consistency of behavior between "set" and "get" APIs. Also, the advanced
share_storage()APIs have been replaced with
- Version 2.19.01 turned off multi-pass tuning by default. Enable with
- Version 2.18.03 allowed the default radius to be stencil-specific and changed the names of example stencil "9axis" to "3axis_with_diags".
- Version 2.18.00 added the ability to specify the global-domain size, and it will calculate the local-domain sizes from it. There is no longer a default local-domain size. Output changed terms "overall-problem" to "global-domain" and "rank-domain" to "local-domain".
- Version 2.17.00 determined the host architecture in
bin/yask.shand number of MPI ranks in
bin/yask.sh. This changed the old behavior of
-ranks. Those options are still available to override the host-based default.
- Version 2.16.03 moved the position of the log-file name to the last column in the CSV output of
- Version 2.15.04 required a call to
yc_grid::set_dynamic_step_alloc(true)to allow changing the allocation in the step (time) dimension at run-time for grid variables created at YASK compile-time.
- Version 2.15.02 required all "misc" indices to be yask-compiler-time constants.
- Version 2.14.05 changed the meaning of temporal sizes so that 0 means never do temporal blocking and 1 allows blocking within a single time-step for multi-pack solutions. The default setting is 0, which keeps the old behavior.
- Version 2.13.06 changed the default behavior of the performance-test utility (
yask.sh) to run trials for a given amount of time instead of a given number of steps. As of version 2.13.08, use the
-trial_timeoption to specify the number of seconds to run. To force a specific number of trials as in previous versions, use the
- Version 2.13.02 required some changes in perf statistics due to step (temporal) conditions. Both text output and
- Version 2.12.00 removed the long-deprecated
==operator for asserting equality between a grid point and an equation. Use
- Version 2.11.01 changed the plain-text format of some of the performance data in the test-utility output. Specifically, some leading spaces were added, SI multipliers for values < 1 were added, and the phrase "time in" no longer appears before each time breakdown. This may affect some user programs that parse the output to collect stats.
- Version 2.10.00 changed the location of temporary files created during the build process. This will not affect most users, although you may need to manually remove old
- Version 2.09.00 changed the location of stencils in the internal DSL from
.cppfiles. See the notes in https://github.com/intel/yask/releases/tag/v2.09.00 if you have any new or modified code in