In the first benchmark test, we examine the performance of computing the eighth number in the Fibonacci sequence. Note that we pass the -bench=. argument, which means run all benchmark tests. The ./... argument means to run all the tests in this directory and all the child directories as well:
When we request the eighth number in the sequence, the simple recursive implementation runs faster than the memoized and channeled (optimized) versions, 213 ns/op compared to 1302 ns/op and 2224 ns/op, respectively.
In fact, when the simple version is executed once, it only takes 3.94 ns/op.
One very cool feature of Go's benchmark ...