[Stackless] Newbie questions

Konrad Hinsen hinsen at cnrs-orleans.fr
Mon Feb 17 19:39:16 CET 2003


On Monday 17 February 2003 17:17, Paul Sheer wrote:

> the big question is "how do you determine the granularity
> of the communicate-compute cycle so that you are not stuck
> back with the same problem as you had with ordinary threads?"

Good question. Personally, I am not much interested in fine-grained stuff as 
neither my hardware nor my typical problems are adapted to it. I have never 
been able to use more than 10 processors for a single production job. BSP 
experts tell me that BSP works just as well for fine-grained parallelism, but 
I have no personal experience.

> can you please go to threading.2038bug.com and tell me if you
> have any comments? i will be adding BSP to the list of links

"Threads are evil" is a bit too strong, in my opinion. I see threads as an 
OS-level tool. Just like hand-coding in assembler, it is sometimes a useful 
technique. But threads are not a good solution to put concurrency into 
applications. Unfortunately, no one has proposed a better universal solution 
yet.

I have used both threading and MPI parallelism in my applications, and my 
experience with both is that I would rather avoid them :-)

> my phd will be to create a compiler that compiles ansi-C code into
> ansi-C code that can be executed as multiple tasklets under the
> same process. it will do this by "unrolling" all the code around

Sounds a bit like "stackless C"  ;-)

Konrad.

_______________________________________________
Stackless mailing list
Stackless at www.tismer.com
http://www.tismer.com/mailman/listinfo/stackless




More information about the Stackless mailing list