Benchmark

[PrimaryKey]

Hello All, I have a question about benchmarking the channels. I didnt find any info about that and because of that I run a small test myself. The test pointed to the channel wires being ~3 times or more slower than queues with the same data. I'm including my benchmark (not my code, kudos to Alexander Glasner ). Please let me know if I'm doing something incorrectly. 

When I dug deeper into the channel VIs i found a level on which the API has debugging enabled. Might this be slowing it down?

[Duffy2007]

Hello to all

the interesting part is, when the channel wires for streaming are in use, they are very near on the results when using a queue, if a channel wire is created new with every call a overhead will be created in the background. Thats the difference bettween the benchmarks normal and inside of a For loop.

Additional thoughts about channel wire messages with ACK internal on a windows machine a DLL or process will be called in an interval from an internal timer inside of windows. The value for the interval can be read out with the Windows System Timer Tool. This could explain this a little bit, but not complete because  the time from the benchmark is no linear to the run tests.