|
||||||||||
PREV PACKAGE NEXT PACKAGE | FRAMES NO FRAMES |
See:
Description
Interface Summary | |
Operation | Interface for implementing arbitrary operations to be executed. |
OperationExecutor | Interface for implementing objects that can execute Operation s.
|
Class Summary | |
ApfloatHolder | Simple JavaBean to hold one Apfloat .
|
BackgroundOperation | Class for running an Operation in the background in a separate thread. |
LocalOperationExecutor | Class to execute Operation s locally.
|
OperationServer | Server for executing Operation s from remote calls. |
Pi | Calculates pi using three different algorithms. |
Pi.BorweinPiCalculator | Calculates pi using the Borweins' quartic algorithm. |
Pi.ChudnovskyPiCalculator | Basic class for calculating pi using the Chudnovskys' binary splitting algorithm. |
Pi.ChudnovskyProgressIndicator | Indicates progress of the pi calculation using the Chudnovskys' binary splitting algorithm. |
Pi.GaussLegendrePiCalculator | Calculates pi using the Gauss-Legendre algorithm. |
PiApplet | Applet for calculating pi using three different algorithms. |
PiDistributed | Calculates pi using a cluster of servers. |
PiDistributed.DistributedPiCalculator | Distributed version of the Chudnovskys' binary splitting algorithm. |
PiParallel | Calculates pi using multiple threads in parallel. |
PiParallel.ParallelPiCalculator | Parallel version of the Chudnovskys' binary splitting algorithm. |
PiParallel.ThreadLimitedOperation | Class to execute operations while setting ApfloatContext.setNumberOfProcessors(int)
to some value. |
PiParallelApplet | Applet for calculating pi using multiple threads in parallel. |
RemoteOperationExecutor | Class to call an OperationServer to execute Operation s remotely. |
Sample applications demonstrating apfloat use.
Three different versions of an application for calculating π are
included. The simplest, Pi
runs on one
computer using one processor (and thread) only. PiParallel
executes multiple threads in parallel and has vastly better performance
on symmetric multiprocessor (SMP) computers. Finally,
PiDistributed
can use multiple separate
computers for calculating pi for even greater calculating power.
As a curiosity, two applets are provided for running Pi
and PiParallel
through a graphical user
interface: PiApplet
and PiParallelApplet
,
correspondingly.
Compared to the C++ version of apfloat, the Java version pi calculation program does fairly well. It achieves roughly 50% of the performance of the assembler-optimized C++ versions of apfloat. But, compared to the plain ANSI C++ version the Java version can be as much as 25% faster! This truely shows that adaptive compilation from an intermediate binary executable format is a better solution than direct compilation to a single native format.
|
||||||||||
PREV PACKAGE NEXT PACKAGE | FRAMES NO FRAMES |