|
Read a recent article on
Fork-Join
Processing by one of our developers about the
issues with
multi-core development and why you need this product. |
What is Fork-Join?
Think of a fork
in the road where each path eventually comes back together —
joins.
Fork-Join breaks
an application into several parts for parallel processing and joins
the results at the end.
Figure
1: Fork-Join Structure
Let’s say
we have an array of one thousand numbers. We need to do a procedure
on each of these numbers and add the total.
Listing
1: Array
Processing
|
for
(int i = 0; i < 1000; i++)
{
total +=
doProcedure(array[i]);
}
|
If the procedure
takes one second (wall-clock time) to complete, then it is going to
take one thousand seconds (over 16½ minutes) to complete this
task.
Fork-Join
could
- separate (fork)
the large array into ten arrays of one hundred elements
each,
- processes each
array on a separate CPU, and
- join the
results when finished.
That would take
one hundred seconds (just over 1½ minutes), one tenth of the
original time. The more CPU's available, the faster the
result.
This abstraction closely
resembles the standard scientific model of Divide-and-Conquer.
Figure
2: Divide-and-Conquer
This is what
high performance computing is all about — simultaneously processing
humongous amounts of data on as many CPU’s as
available. Map/Reduce, sorting, and countless array processing applications all
can use Divide-and-Conquer to solve their problems.
And there's more. Lots, lots more, so keep
reading.
Would you like to segregate
processing into separate Java™
Virtual Machines without the
gargantuan overhead of commercial application
servers?
Do you need many flavors of RMI Servers
from
the simple to the complex?
Would you like to run a backend queuing and
threading server embedded
within any application?
Is it too time-consuming to develop a custom server that
can handle
Timed, Autonomous, Urgent and Multi-Component
requests?
If you need a secure,
reliable, manageable and fault tolerant Divide-and-Conquer
Server
for any purpose and do
not wish to start at the
beginning designing and testing one yourself,
then you need Tymeac™
