That's a fairly vague question. I use thread, data_time, chrono, filesystem, and system on an older embedded ARM processor. These libraries are great for my application. I recommend that you not just grab anything you see in Boost indiscriminately, but rather (as with anything else) look a little deeper and make sure the specific library is suitable. I think it's fair to say that all Boost libraries have efficiency as a consideration, but are most uncompromising about correctness. A few years back I used Boost.Python and found that it generated surprisingly large code in order to be correct in cases that didn't matter to me. The code size almost became problematic. Some libraries (I don't have an example at hand) do things that pretty clearly imply a fair amount of malloc() and free(). Just because it's Boost doesn't mean it's going to be suitable for your embedded application. Steven J. Clark VGo Communications From: Boost-users [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Trek Sent: Thursday, April 23, 2015 5:04 AM To: boost-users Subject: [Boost-users] boost on embedded target such as ARM how is boost doing on embedded environment in terms of efficiency? thanks