Skip to content
GitLab
Projects Groups Topics Snippets
  • /
  • Help
    • Help
    • Support
    • Community forum
    • Submit feedback
  • Register
  • Sign in
  • E ear_learning
  • Project information
    • Project information
    • Activity
    • Labels
    • Members
  • Repository
    • Repository
    • Files
    • Commits
    • Branches
    • Tags
    • Contributor statistics
    • Graph
    • Compare revisions
  • Issues 0
    • Issues 0
    • List
    • Boards
    • Service Desk
    • Milestones
  • Merge requests 0
    • Merge requests 0
  • CI/CD
    • CI/CD
    • Pipelines
    • Jobs
    • Schedules
  • Deployments
    • Deployments
    • Environments
    • Releases
  • Packages and registries
    • Packages and registries
    • Package Registry
    • Container Registry
    • Terraform modules
  • Monitor
    • Monitor
    • Incidents
  • Analytics
    • Analytics
    • Value stream
    • CI/CD
    • Repository
  • Wiki
    • Wiki
  • Snippets
    • Snippets
  • Activity
  • Graph
  • Create a new issue
  • Jobs
  • Commits
  • Issue Boards
Collapse sidebar
  • EAR_teamEAR_team
  • ear_learning
  • Wiki
  • 1 ˗ Kernels compile

1 ˗ Kernels compile · Changes

Page history
Updated wiki. authored May 27, 2019 by xgomez's avatar xgomez
Show whitespace changes
Inline Side-by-side
1-˗-Kernels-compile.md 0 → 100644
View page @ a008fe08
Setting up
----------
The compiling script is located in `scripts/learning/kernels_compile.sh`. Before execute it, you have to perform some adjustments:
1) Open `kernels_compile.sh` and look for these lines:
```
# Edit architecture values
export CORES=28
export SOCKETS=2
export CORES_PER_SOCKET=14
```
2) Update the following parameters:<br />
- **CORES**: the total number of cores in a single computing node.<br />
- **SOCKETS**: the total number of sockets in a single computing node.<br />
- **CORES_PER_SOCKET**: the total number of cores per socket in a single computing node.<br />
Kernel compilation
------------------
The first step is the kernel compilation. You can execute the script in `script/learning/kernels_compile.sh` and the compiled kernels will be saved in the `bin` folder in the root of this package. If something goes wrong, you can surf in `kernels` folder and edit the make files to switch the `mpiifort` compiler or any other option for your own, because the compiling script it does nothing but call that make commands.
You can also compile one by one manually. In that case, open the script `scripts/learning/helpers/kernels_executor.sh` and look at the function `learning()`. Here you will find all the paths and compile instructions (make) and install (move) for each kernel when `$BENCH_MODE` variable is set to `compile`.
Additional kernel information
-----------------------------
You can see some information about all the kernels in the following table:
| Kernel | Reference |
| ------ | -------------------------------------------------------------------------------- |
| dgemm | https://software.intel.com/en-us/mkl-tutorial-c-multiplying-matrices-using-dgemm |
| stream | http://www.cs.virginia.edu/stream/ref.html |
| bt-mz | https://www.nas.nasa.gov/publications/npb.html |
| sp-mz | " |
| lu-mz | " |
| ep | " |
| lu | " |
| ua | " |
\ No newline at end of file
Clone repository
  • 1 ˗ Kernels compile
  • 2 ˗ Nodes lists guidelines
  • 3 ˗ Kernels test
  • 4 ˗ Kernels launch
  • 5 ˗ Coefficients compute
  • Home