Revamp intro to multi-locale execution documentation

[ronawho]

Our current multi-locale execution docs are geared towards how to run distributed across multiple nodes, but most users (especially those just getting started) will be running/debugging on a single node that's emulating multiple locales. We should make that common case easier to find.

[ronawho]

udp defaults to segment everything, which doesn't support pshm. We might want to recommend smp (which defaults to segment fast and has pshm support) since that can be much faster -- https://gitter.im/chapel-lang/chapel?at=5cc38cdd97dcb371d81cabd7

pshm info:

      This feature uses shared memory communication among the processes (aka
      GASNet nodes) within a single compute node (where the other alternatives
      are multi-threading via a PAR or PARSYNC build; or use of the conduit's
      API to perform the communication).
      Note that not all conduits and operating systems support this feature.
      For more information, see the section below entitled "GASNet inter-Process
      SHared Memory (PSHM)".

[LouisJenkinsCS]

Definitely need to make CHPL_GASNET_SEGMENT=fast the default, its a night and day difference!

NumLocales	SMP/UDP-Fast	UDP-Everything (Default)
1	0.009726	009726
2	0.016445	0.570676
4	0.025065	0.984162
8	0.051521	1.2397
16	0.099913	3.57597
32	0.17136	7.76913
64	0.261266	11.4257