[mpich-discuss] MPICH-enabled JWM?

[Eugene N]

Hello

Well thats a known fact - interpreted languages suck at performance. Not to
mention that
multithreading on certain OS's can be real slow.

GCC can compile java into native object code, perhaps then it would made a
difference?

the reason for me asking about distributed JWm was trivial, a friend of mine
has two old puters and wants to play minecraft :P

2011/3/7 Mandar Gurav <mandarwce at gmail.com>

> Recently, we wrote some compute intensive sample programs both using
> C(openmp) and java(Multithreading). And found that the Java programs
> took about twice of the time for execution than C. And as compared to
> this kind of worst performance the openmp and MPI seem to have good
> performance.
>
> On Sat, Mar 5, 2011 at 3:37 AM, Dmitri Chubarov
> <dmitri.chubarov at gmail.com> wrote:
> > Hi, Eugene,
> >
> > there are projects that might fit your description.
> >
> > There are many research projects that can be found in the literature. For
> > instance there was a distributed memory JVM developed by Henri Bal in
> > Amsterdam [1] That project's aim was to achieve good performance
> > characteristics and that required compiler support. For a pure bytecode
> > execution references to several SSI Java Machines can be found [2]. One
> such
> > projects provides code for download [3].
> >
> > Though I do not have any references neither on the use of distributed jvm
> in
> > production nor on implementations of distributed java machines in MPI.
> >
> > Dima
> >
> > [1] R. Veldema, R. F. H. Hofman, R. A. F. Bhoedjang, and H. E. Bal.
> Runtime
> > Optimizations for a Java DSM Implementation. In 2001 joint ACM-ISCOPE
> Conf.
> > on Java Grande, pages 153–162, Palo Alto, CA., June 2001.
> >
> > [2] C.Cavanna et al. Jupiter/SVM: A JVM-based Single System Image for
> > Clusters of Workstations, In Proc. 16th Parallel and Distributed
> Computing
> > and Systems, 2004.
> >
> > [3] dJVM, http://cs.anu.edu.au/djvm/download.php
> >
> >
> >
> > On Sat, Mar 5, 2011 at 3:10 PM, Eugene N <neverov.biks.07.1 at gmail.com>
> > wrote:
> >>
> >> Hello
> >>
> >> I wonder, is there a mpi-based JWM, so one could run bloated java
> >> "programs" on a cluster of normal workstations??
> >>
> >>
> >> Eugene
> >>
> >> _______________________________________________
> >> mpich-discuss mailing list
> >> mpich-discuss at mcs.anl.gov
> >> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
> >>
> >
> >
> > _______________________________________________
> > mpich-discuss mailing list
> > mpich-discuss at mcs.anl.gov
> > https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
> >
> >
>
>
>
> --
> Mandar Gurav
> http://www.mandargurav.org
> _______________________________________________
> mpich-discuss mailing list
> mpich-discuss at mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/mpich-discuss/attachments/20110307/3b883572/attachment.htm>