I was asked an interesting question regarding my blog post on the tests that compared ADO and ADO.NET from VBscript and PowerShell: Was it using native PowerShell ISE or "cold start" execution?
The answer: Native shell execution
I ran the VBscript code in a standard CMD shell. I ran both of the PowerShell code examples in the PowerShell ISE shell.
When I ran each script using a "cold start" process, it added a half-second to VBscript and 1.5 seconds to the PowerShell tests. That tells me that the PowerShell "engine" is slower to initialize from a cold launch request, which could be a combination of many factors from .exe size, thread starts, API requests, and so on.