Most modern threading platforms are already offering task based programming models. Thus, they are allowing developers to follow one of the eight rules for multicore programming written by James Reinders a few years ago. I’m specifically talking about rule #3: "Program in tasks (chores), not threads (cores)."
Parallelism get much more fun when you have enough of it... today we got to see a vision of the future with very impressive demos: The "Single-chip Cloud Computer" (SCC) - was first shown publicly today and is destined for usage in research inside and outside Intel!
Justin Rattner announced a new research milestone in manycore computing: the most Intel architecture (IA, aka x86) processors on a single piece of silicon ever. It has 48 cores hooked together in network that mimics cloud computing on a chip level, and support highly parallel "scale-out" programming models.
Intel® Parallel Studio Service Pack 1 is now available, adding support for Windows* 7.
SP1 is well worth downloading and installing - here are some of the reasons:
- Parallel Inspector and Parallel Amplifier can be driven (for automating test suites) from the command line now.
- Bug fixes - of course - not many issues needed fixing, but you may appreciate the ones bugs that were found and fixed!
More information about "Ct technology" and the ability to apply for the beta program is at intel.com/software/data_parallel
It has been relatively easy for us to follow the path of a serial application. The flow of the application is managed by the Stack. This makes the execution flow of a single process to be managed by CPU hardware. Today we face the need to execute several processes in parallel and thus have several execution paths at the same time. This is harder for us to manage and keeping track of this flow is complex. The hardware accelerated Stack is no longer sufficient for us and the Stack-Trace is no longer the state of the application.