Intel® Threading Building Blocks

Cuncurrent Containers vs STL Containers + Mutex/Lock

I am wondering, ifthe efficience issue is the ONLY factor concerned,

could we replace the Intel Official TBB Cuncurrent Containers with STL Containers in cooperation with TBB mutex?

If not, please specify the by efficiencedifference with some testing dataon timing.

If so,how to choose thebestkind of mutexes fora particular circumstance? like, to perform push/pop/sort/swap/... ona queue/vector?

problems with TBB app


Ive downloaded TBB (tbb21_20080605oss) from this site, made simple application in Microsoft Visual Studio 2008 Express based on tree_sum sample.It works fine on my PC (Vista SP1/Core 2 Quad 2.4 Ghz).But when Im trying to start it on other computers (such as XP SP2/Pentium D 3 Ghz) I get an could not launch the application error.

Depends.exe shows that all needed dlls are present (tbb.dll, tbbmalloc.dll, msvcp90.dll, msvcr90.dll).

What could be the cause?

is TBB parallel-for work with SSE2 subroutine ?

As I know, doing parallel for is just decomposing for-action into several threads.

My question is, can I run SSE2-coded subroutines in parallel ?

A dual processor has only two sets of 128bit SSE2 registers. Does it mean I could only run two threads at a time ?

I have written a simple parallel-for with SSE2 function embedded. So far, the program run nicely.

I wonder if my code is unsafe(race condition problem), since the number of SSE registers are limited.

Problems on concurrent_vector with TBB 2.1 by VisualStudio2005sp1 under WinXP SP2

I am currently usingTBB 2.1inVisualStudio2005sp1 under WinXP SP2, with the Intel official VS plug-in for TBB.

I am quite sureI has installed the TBBlibrary and the VS plug-in properly, and I have written some codes with TBB[not containing the concurrent_vector] that compiled and run successfullyin VS. And here is a special case I cannot solve nor can I find any reference from the turtorial or manual. It is playing withpointers to the concurrent_vector, compiled with debug version.

Forums will be read-only from 09/24 10PM PST to 09/25 12PM PST


From Tuesday, September 24, 10pm PST to Thursday, September 25, 12pm PST, the Intel Software Network Community Forums will be read only in preparation for the new forums interface launch on Thursday, September 25. During this time, users will be able to read the forums, but will not be able to reply or submit new topics. The new forums will have an improved look and many added features. We hope youll like the new ISN Community Forums.

Warm regards,

cache_aligned_allocator and small allocations

ive got a cache_aligned_allocator which I use to allocate matrices that are at most 16bytes...

now im unsure how cache_aligned_allocator works with allocations smaller than a cache line...

does it allocate 16 bytes and then add unusuable padding?

or does it allocate as many matrices there is room for in the cacheline and then pads the rest... and then when i allocate another matrix it will actually use one of these "padding matrices"...



TBB uses an approximation of Cilk-style
scheduling (work LIFO; steal FIFO).It has good locality and load
balancing properties while making certain space guarantees. It enables
clean nesting of software components, analogous to the way serial code
composes by nested subroutine calls.

Подписаться на Intel® Threading Building Blocks