-
Notifications
You must be signed in to change notification settings - Fork 17
Home
Two moderated mailing-lists are available, one for the users and the other for the developers involved in the project. The user mailing list subscriptions are handled via the user mailing-list, while the developers via the developers mailing-list. Please feel free to join us.
We use DPLASMA as a demonstrator on how to compile PaRSEC and use it to compile PaRSEC enabled operations.
-
Understanding the compilation and runtime process of PaRSEC
- How to write a PaRSEC enabled operation by using the PaRSEC compiler
- How to write a PaRSEC enabled operation by editing the JDF file directly
- How to write a program that uses a PaRSEC enabled operation (a.k.a. "what is this data distribution thing?")
- How to profile PaRSEC applications
- How to write a PINS Module to instrument PaRSEC applications
- How to use the Python PaRSEC Trace Tables interface
- How to run PaRSEC in Intel Xeon Phi (native mode)
- Visualization: Stream algorithmic properties out of the PaRSEC runtime
- Binding with MPI and PaRSEC
- Migrating Legacy Applications to PaRSEC
The PaRSEC / DPLASMA community has put together tutorials to explain how to use PaRSEC in a distributed heterogeneous setting, how to implement your own Domain Specific Language on top of PaRSEC, and how to improve the efficiency of your algorithms by using a dataflow-like task-based programming environment. These tutorials are targetted toward the user community, but potential developers of PaRSEC might find some interesting information as well. Tutorial 1 and Tutorial 2.
- Memo on HPL over PaRSEC
- Memory errors detectors
- A high level description of the PaRSEC runtime
- Memo on proposal for Collective Design
- Tips/Experiences learning JDF from a graduate student
----
Hints on the wiki editing (the previous welcome page)