On Tue, May 18, 2010 at 2:00 PM, Cedric Pradalier
<
cedric.pradalier@mavt.ethz.ch> wrote:
> Hi,
>
> A couple of week ago, I've posted a pdf with performance evaluation for
> inter-node only using local TCP or Sharedmemory. I measured message
> delivery time, and message frequency, as a function of message size.
> Search for the "Shared memory image plugin" thread.
>
> Obviously, these transports cannot compete with the nodelet architecture...
>
Hello, I'd like to just shine a light on this last statement. I don't
immediately understand why a shared memory transport can't compete
with nodelets as I understand them. If nodelets are threads in a
single process and use process memory to carry messages (which is
essentially shared memory in that context), shouldn't nodelets and a
shared memory transport be roughly equivalent in performance?
thanks,
-andrew
(text/plain)