The eleventh month of work focused on work in the following areas:
- PhyLS Compiler Refactoring
- Performance analysis
- HPXMP – HPX OpenMP backend
- Distributed execution
No comments
Category: Phylanx
Phylanx Report: June 2018
The tenth month of work focused on work in the following areas:
- OTF2 Improvements
- Code view visualization
- Initial implementations of BFS, Neural Network, and LU decomposition
Phylanx Report: May 2018
The Ninth month of work focused on work in the following areas:
- Implementing Solvers
- Work on HPXMP for OpenBLAS parallelization support
- Bug fixes in APEX
- Visualization
- HPX-Python interactions
Phylanx Seminar: The Primitive Edition
A core abstraction in the Phylanx software stack is the primitive. Primitives define the operations which are needed to execute a user’s code such as addition, subtraction, multiplication, etc. While our team has Continue reading
Phylanx Year 2 Kickoff Meeting
On Thursday and Friday (April 19th and 20th) the Phylanx team meet in Baton Rouge to organize the second phase of the project. At the meeting, we provided updates on the state of the project as well as a Continue reading
Phylanx Report: March 2018
The seventh month of work focused on work in the following areas:
- Performance analysis
- Initial performance comparisons of Phylanx to Numpy, Dask, SciKit Learn, and TensorFlow
- Performance analysis of ALS
- APEX
- Adding support for thread parent tracking
- Visualization
- HPX version 1.1.0 Release
Phylanx Seminar: Revealing the Magic of Blaze
Over the past several weeks, many team members have been asking how aspects of the Phylanx project are implemented. In this seminar, Hartmut explains the techniques used by Blaze to optimize matrix operations. He implements a matrix addition example which uses types, templates, and curiously recurring template patterns (CRTP) to reduce the number of temporaries made during the execution of the code. By avoiding these extra allocations in a compiler friendly way, Blaze can drastically reduce the amount of time it takes perform matrix operations. You can find links to the seminar materials below:
Seminar Video: https://www.youtube.com/watch?v=F5E8cOqHmRU
Naive Matrix Implementation: http://stellar.cct.lsu.edu/files/phylanx_seminars/03.21.18_seminar_blaze_magic/naive_matrix_03.21.18.cpp
Optimized Matrix Implementation: http://stellar.cct.lsu.edu/files/phylanx_seminars/03.21.18_seminar_blaze_magic/optimized_matrix_03.21.18.cpp
Phylanx Seminar: Polyhedral Optimizations
The polyhedral model is a mathematics framework used to optimize loop execution. Its use is supported by OpenScop, a specification for storing gathered information of an analyzed code. In this lecture, Rod Continue reading
Phylanx Seminar: PhyFun
On February 23rd, Steve presented a Phylanx seminar on the current state of the Python front-end to Phylanx, PhyFun. This method of binding Phylanx primitives utilizes Python decorators. The Continue reading
Phylanx Seminar: PhySL
On February 9th, Hartmut presented another Phylanx seminar on the current state of the project. In addition to laying out the project in more detail, he described the special language PhySL which he and Continue reading