about parfor and spmd speedup

4 ビュー (過去 30 日間)
Xiaochun
Xiaochun 2011 年 12 月 30 日
hi,
I have tested the speedup of parfor and spmd as
1)for
tic
for n=1:1000
A=B*C;
end
toc;
2)parfor spmd in one worker in the same pc:
// I have stored B C in the worker before computing
tic
parfor n=1:1000
A=B*C;
end
toc;
I find that the speed of parfor and spmd is only 2/3 of single matlab. is it possible to make it faster?
  1 件のコメント
Walter Roberson
Walter Roberson 2011 年 12 月 30 日
how big are B and C ?

サインインしてコメントする。

採用された回答

Walter Roberson
Walter Roberson 2011 年 12 月 30 日
If your matrices are not very big, then the time required to set up workers and communicate data in to them can overwhelm any saving of using multiple workers.
  3 件のコメント
Walter Roberson
Walter Roberson 2011 年 12 月 30 日
With a matrix that big, your "for" matrix multiplication would be farmed out to the highly optimized and multithreaded BLAS or LAPACK, using multiple cores or hyperthreading.
I gather that each "parfor" or "smpd" worker requires a new process, and my understanding is that those would only have access to the core or hyperthread they are allocated to.
Xiaochun
Xiaochun 2012 年 1 月 2 日
I see. Thank you very much.

サインインしてコメントする。

その他の回答 (1 件)

Knut
Knut 2012 年 1 月 3 日
So the spot where parfor can really shine is loops that are : a)Simple enough that parfor can do them in parallell b)complex/non-general enough that MATLAB does not have a low-level library that does it in paralell anyways
?
  1 件のコメント
Titus Edelhofer
Titus Edelhofer 2012 年 1 月 3 日
As a rule of thumb: yes. Another way to formulate b): when you have e.g. a quad core (with hyperthreading), i.e., your task manager shows you 8 CPU usages, and your (loop) code runs at exactly 12.5% average CPU usage ...

サインインしてコメントする。

カテゴリ

Help Center および File ExchangeMATLAB Parallel Server についてさらに検索

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by