Databases
StackOverflow2013
Postsdbo
POEigen::Matrix vs. boost::multi_array vs. Eigen::Map
Body
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POEigen::Matrix vs. boost::multi_array vs. Eigen::Map
    text
    copied!<p>I'm getting puzzling results while doing fairly simple tasks to compare the performance of:</p> <ul> <li>Eigen::Matrix</li> <li>boost::multi_array</li> <li>boost::multi_array mapped to Eigen::Matrix using Eigen::Map</li> </ul> <p>This is an abridged version of my test code; a fuller version can be found at: <a href="http://pastebin.com/faZ7TvJG" rel="nofollow">http://pastebin.com/faZ7TvJG</a>.</p> <pre><code>boost::multi_array&lt;double, 2, Eigen::aligned_allocator&lt;double&gt; &gt; boost_multi_array; Eigen::Matrix&lt;double, Eigen::Dynamic, Eigen::Dynamic, Eigen::RowMajor&gt; eigen_matrix; Eigen::Map&lt;Eigen::Matrix&lt;double, Eigen::Dynamic, Eigen::Dynamic, Eigen::RowMajor&gt; &gt; boost_multi_array_mapped(boost_multi_array.data(), rows, cols); double tmp_sum = 0, tmp_time = omp_get_wtime(); for(size_t i=0; i&lt;iterations; i++) { for(size_t j=0; j&lt;rows; j++) { for(size_t k=0; k&lt;cols; k++) { //if(k%2==0) //{ // commented out are the different options //tmp_sum += boost_multi_array[j][k]; //tmp_sum += boost_multi_array_mapped(j,k); tmp_sum += eigen_matrix(j,k); //} } } } const double sequential_access_time = omp_get_wtime() - tmp_time; </code></pre> <p>The results are as follows:</p> <pre><code>Sequential Access: BOOST (MAPPED) : 1.45763s EIGEN : 1.45736s BOOST : 2.58971s </code></pre> <p>If I use an if-statement to skip every second element, I then get the following results:</p> <pre><code>Alternating Access: BOOST (MAPPED) : 1.67301s EIGEN : 2.08834s BOOST : 2.35295s </code></pre> <p>Inspecting the assembly shows that in the sequential access case, Eigen is faster because the sum becomes vectorized, while it does not when using raw boost::multi_array.</p> <p>My questions then are:</p> <ol> <li>Why is boost::multi_array not vectorized, while Eigen::Matrix is?</li> <li>Why would a multi_array mapped to Eigen be faster than a "native" Eigen data structure?</li> </ol> <p>For compilation I use the following:</p> <pre><code>g++ -I /usr/include/eigen3 test.cpp -Wall -O3 -DNDEBUG -DBOOST_DISABLE_ASSERTS -fopenmp -ffast-math -funroll-loops -march=native -mtune=native -o array_test </code></pre> <p>Thanks for your answers.</p>
 

Querying!

 
Guidance
An individual column

Larger individual text columns get their own page to allow for proper reading.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload