Control Replication: Compiling Implicit Parallelism to
Efficient SPMD with Logical Regions
SessionCompilation Techniques
Authors
Event Type
Paper
Compiler Analysis and Optimization
Parallel Programming Languages, Libraries, Models
and Notations
Program Transformation
Programming Systems
TimeTuesday, November 14th4pm -
4:30pm
Location402-403-404
DescriptionWe present control replication, a technique for
generating high-performance and scalable SPMD code from
implicitly parallel programs. In contrast to traditional
parallel programming models that require the programmer
to explicitly manage threads and the communication and
synchronization between them, implicitly parallel
programs have sequential execution semantics and by
their nature avoid the pitfalls of explicitly parallel
programming. However, without optimizations to
distribute control overhead, scalability is often
poor.
Performance on distributed-memory machines is especially sensitive to communication and synchronization in the program. Thus optimizations for these machines require an intimate understanding of a program's memory accesses. Control replication achieves particularly effective and predictable results by leveraging language support for first-class data partitioning in the source programming model. We evaluate an implementation of control replication for Regent and show that it achieves up to 99% parallel efficiency at 1024 nodes with absolute performance comparable to hand-written MPI(+X) codes.
Performance on distributed-memory machines is especially sensitive to communication and synchronization in the program. Thus optimizations for these machines require an intimate understanding of a program's memory accesses. Control replication achieves particularly effective and predictable results by leveraging language support for first-class data partitioning in the source programming model. We evaluate an implementation of control replication for Regent and show that it achieves up to 99% parallel efficiency at 1024 nodes with absolute performance comparable to hand-written MPI(+X) codes.
Download PDF:
here
Authors
