new multicore programming docs for GCC and Visual Studio

Le
cynko
http://www.cilk.com/resources-for-multicoders/for-developers-only/

documentation, code samples, training materials for developers

Cilk++ ships in several months, but the above is a preview of what's coming.
Vidéos High-Tech et Jeu Vidéo
Téléchargements
Vos réponses Page 1 / 2
Gagnez chaque mois un abonnement Premium avec GNT : Inscrivez-vous !
Trier par : date / pertinence
Dmitriy V'jukov
Le #16794261
On Sep 12, 3:09 pm, "cynko"
http://www.cilk.com/resources-for-multicoders/for-developers-only/

documentation, code samples, training materials for developers

Cilk++ ships in several months, but the above is a preview of what's comi ng.



It seems that you meant '*HPC* multicore programming'. Do you have
anything for development of smart clients, or servers, systems
programming, middleware etc?

Dmitriy V'jukov
cynko
Le #17248301
from the website - I think this is broader than traditional HPC:

"Cilk++ is best suited for applications that meet the following criteria:

1.. Application performance is compute bound
2.. Performance can be improved by accelerating serial (i.e.,
single-threaded) portions of the application"

"Dmitriy V'jukov" news:
On Sep 12, 3:09 pm, "cynko"
http://www.cilk.com/resources-for-multicoders/for-developers-only/

documentation, code samples, training materials for developers

Cilk++ ships in several months, but the above is a preview of what's
coming.



It seems that you meant '*HPC* multicore programming'. Do you have
anything for development of smart clients, or servers, systems
programming, middleware etc?

Dmitriy V'jukov
Chris M. Thomasson
Le #17271391
"cynko" news:
http://www.cilk.com/resources-for-multicoders/for-developers-only/

documentation, code samples, training materials for developers

Cilk++ ships in several months, but the above is a preview of what's
coming.



What work-stealing algorithm are you using?
gremlin
Le #17365411
"Chris M. Thomasson" news:07KAk.13389$

What work-stealing algorithm are you using?



http://en.wikipedia.org/wiki/Cilk#Work-stealing
Chris M. Thomasson
Le #17367861
"gremlin" news:

"Chris M. Thomasson" news:07KAk.13389$

What work-stealing algorithm are you using?



http://en.wikipedia.org/wiki/Cilk#Work-stealing



I am missing something here. Can you point me to the actual algorithm for
the atomic deques? There are several algorihtms out there, but I am
interested in which one Clik uses. Is it the ABP algorithm? Which one is it?
Thanks for your time and help.
Chris M. Thomasson
Le #17367851
"Chris M. Thomasson" news:FsSDk.4177$
"gremlin" news:

"Chris M. Thomasson" news:07KAk.13389$

What work-stealing algorithm are you using?



http://en.wikipedia.org/wiki/Cilk#Work-stealing



I am missing something here. Can you point me to the actual algorithm for
the atomic deques? There are several algorihtms out there, but I am
interested in which one Clik uses. Is it the ABP algorithm? Which one is
it? Thanks for your time and help.



Are you using something like the algorithm in TBB?
gremlin
Le #17371251
"Chris M. Thomasson" news:JuSDk.4178$

I am missing something here. Can you point me to the actual algorithm for
the atomic deques? There are several algorihtms out there, but I am
interested in which one Clik uses. Is it the ABP algorithm? Which one is
it? Thanks for your time and help.



Are you using something like the algorithm in TBB?



http://publications.csail.mit.edu/lcs/pubs/pdf/MIT-LCS-TM-548.pdf

TBB is based on MIT's Cilk work:
http://www.cilk.com/multicore-blog/bid/5607/Cilk-Wins-Most-Influential-PLDI-Paper-Award
Chris M. Thomasson
Le #17371771
"gremlin" news:

"Chris M. Thomasson" news:JuSDk.4178$

I am missing something here. Can you point me to the actual algorithm
for the atomic deques? There are several algorihtms out there, but I am
interested in which one Clik uses. Is it the ABP algorithm? Which one is
it? Thanks for your time and help.



Are you using something like the algorithm in TBB?



http://publications.csail.mit.edu/lcs/pubs/pdf/MIT-LCS-TM-548.pdf

TBB is based on MIT's Cilk work:
http://www.cilk.com/multicore-blog/bid/5607/Cilk-Wins-Most-Influential-PLDI-Paper-Award



How does the internal low-level impl compare to the following:

http://www.cs.bgu.ac.il/~hendlerd/papers/dynamic-size-deque.pdf

I am looking for raw pseudo-code for atomic deque internal impl details...
AFAICT, this work from SUN would scale better than Clik. Please correct me
if I am way off base here. It seems like spawning a successor thread has
overheads... Humm. Pleas try to bear with me here; okay? Correct my
ignorance on Clik's work-stealing internal impl... Well, let me pick an impl
to focus on... Say, DEC Alpha?
Chris M. Thomasson
Le #17371761
"Chris M. Thomasson" news:BG3Ek.354$
"gremlin" news:

"Chris M. Thomasson" news:JuSDk.4178$

I am missing something here. Can you point me to the actual algorithm
for the atomic deques? There are several algorihtms out there, but I am
interested in which one Clik uses. Is it the ABP algorithm? Which one
is it? Thanks for your time and help.



Are you using something like the algorithm in TBB?



http://publications.csail.mit.edu/lcs/pubs/pdf/MIT-LCS-TM-548.pdf

TBB is based on MIT's Cilk work:
http://www.cilk.com/multicore-blog/bid/5607/Cilk-Wins-Most-Influential-PLDI-Paper-Award



How does the internal low-level impl compare to the following:

http://www.cs.bgu.ac.il/~hendlerd/papers/dynamic-size-deque.pdf

I am looking for raw pseudo-code for atomic deque internal impl details...
AFAICT, this work from SUN would scale better than Clik. Please correct me
if I am way off base here. It seems like spawning a successor thread has
overheads... Humm. Pleas try to bear with me here; okay? Correct my
ignorance on Clik's work-stealing internal impl... Well, let me pick an
impl to focus on... Say, DEC Alpha?



BTW, here is my "fairly" detailed analysis on the inner-workings of SUN's
atomic deque algorithm:

http://groups.google.com/group/comp.programming.threads/browse_frm/thread/8ad297f61b369a41
Dmitriy V'jukov
Le #17371881
On Sep 29, 4:23 pm, "Chris M. Thomasson"
>> Are you using something like the algorithm in TBB?

>http://publications.csail.mit.edu/lcs/pubs/pdf/MIT-LCS-TM-548.pdf

> TBB is based on MIT's Cilk work:
>http://www.cilk.com/multicore-blog/bid/5607/Cilk-Wins-Most-Influentia...

How does the internal low-level impl compare to the following:

http://www.cs.bgu.ac.il/~hendlerd/papers/dynamic-size-deque.pdf

I am looking for raw pseudo-code for atomic deque internal impl details.. .
AFAICT, this work from SUN would scale better than Clik. Please correct m e
if I am way off base here. It seems like spawning a successor thread has
overheads... Humm. Pleas try to bear with me here; okay? Correct my
ignorance on Clik's work-stealing internal impl... Well, let me pick an i mpl
to focus on... Say, DEC Alpha?




AFAIK, in early days Cilk's work-stealing deque used mutex-based
pop(). But I remember there was some mentions of non-blocking
algorithms in the Cilk's papers, something like "some people point us
that it's possible to implement work-stealing deque in completely non-
blocking manner". And I don't know whether non-blocking deque was
finally incorporated into Cilk.

If mutex is spin-mutex (i.e. there is only 1 atomic RMW per lock/
unlock) and stealing is rare, then mutex-based deque is nearly the
same as non-blocking deque with 1 RMW... provided that push() doesn't
use mutex. And provided that atomic RMW has the same cost as StoreLoad
memory fence (x86).


Dmitriy V'jukov
Publicité
Poster une réponse
Anonyme