Dear Open MPI developers,<div><br></div><div>I&#39;m a graduate student and recently test some minibenchmarks using Open MPI library on Xeon X5650 cluster. From the website of Open MPI, I know that Open MPI can use <p style="display:inline!important">
<b><code>sm</code> BTL</b> to transfer data with high bandwidth via shared memory. My question is that what about  Open MPI collectives on shared memory? Were they implemented and optimized on top of point-to-point communication or utilizing shared memory separately?</p>
</div><div><p style="display:inline!important"><br></p></div><div><p style="display:inline!important">Best Regards,</p></div><div><p style="display:inline!important">Shigang Li.</p></div>