I am currently migrating my applications from using PISDK to the AFSDK as per the recommended direction.
My application is reading COMTrade data from a high speed data recorder. This often produces 32 data channels with 48000 values per channel within 500 ms.
My application reads the data file and writes the data into PI tags.
When I switched to using AFSDK only, I used the call
This call takes 88.7 secs to write 48000 values into PI.
The original PISDK call I used was
This call takes 0.661 secs to write 48000 values to PI.
Given 30,000+ files to process, the AFSDK methods just isn't fast enough.
Is there an alternative, faster method ?