Also, can you confirm that subscribing individually to items with the SubscribeItem() procedure ends to the same result as subscribing to multiple items at once with the SubscribeMultipleItems() procedure?

Hello,

As I'm currently testing out OPCLabs we're facing a limitation in a situation that I may encounter with my application.

I've implemented a TEasyDAClient that subscribe to 1500 tags with a RequestedUpdateRate of 1000ms (1s).
These tags are simulation tags set up to change to a random value every second. This leads to our application receiving 1500 notifications per seconds (this is the stress test I wanted).

As I instantiated the TEasyDAClient in the main thread of my application I'm observing that my application cannot process these OnItemChanged() events quick enough to keep up. Depending on the computer it's running on, it goes from 250 to 550 max processed events per seconds. I understand that there has to be a limit to what can be done on a single core of my processor.

250 operations per seconds should be enough for our current usage of the application but we might hit that ceiling in a near future.
My question is, what do you recommend in terms of implementation if we want to process over 500 operations per second.

Should I implement N threads, each subscribing to X different items?

Thank you for your help.