Dynamic Load balancing: Chare
migration is supported in this release. Migration-based dynamic load balancing
framework with various load balancing strategies library has been added.
Chare Arrays: You can now create an array
of Chare objects and use array index to refer the Charm++ array elements.
A reduction library on top of Chare array has been implemented and included.
Projections: Projections, a Java application
for Charm++ program performance analysis and visualization, has been
included and distributed in the new release. Two trace modes are available:
trace-projections and trace-summary. Trace-summary is a light-weight trace
library compared to trace-projections.
AMPI: AMPI is a load-balancing based library
for porting legacy MPI applications to Charm++. With few changes
in the original MPI code to AMPI, the new legacy MPI application on Charm++
will gain from Charm++'s adptive load balancing ability.
Easy running: Charm++ is making running
Charm++ program easier in this new release. "Charmrun" now available on
all platforms with uniform command line syntax. You can forget the difference
between net-* versions and MPI versions, and run charm++ application with
this same charmrun command syntax. ++local option is added in charmrun
for net-* version, it provides simple local use of Charm and no longer
require the ability to "rsh localhost" or a nodelist file in order
to run charm only on the local machine. This is especially attractive when
you run Charm++ on Windows.
New libraries: Many new libraries have
been added in this release. They include:
master-slave library: for writing manager-worker
paradigm programs.
receiver library: provide asynchronous communication
mode for chare array.
f90charm: provides Fortran90 bindings
for Charm++ Array.
BlueGene: a Charm++/Converse emulator
for IBM's proposed BlueGene machine.
