Virtual PLC Performance
Performance matters because control logic must finish with enough time left for I/O, communications, monitoring, and diagnostics. We publish both the results and the material needed to check them.
Control Seat and CODESYS on the same FR202
We ran ten compiled IEC workloads on one OnLogic FR202, stopping one runtime while the other was measured. Control Seat was faster on nine of the ten workloads, including both PI-style control-loop tests.
The comparison measures how quickly each engine executes the same control logic. It does not measure physical I/O response, network performance, cycle start timing, motion control, or safety response.
What to measure for your application
A benchmark cannot replace commissioning. For the system you intend to run, record:
- controller model and software version;
- application workload and task period;
- test duration and normal background workload;
- p99.9 and maximum scan time;
- p99.9 and maximum cycle-start delay; and
- every overrun or missed cycle.
Long-duration timing results from the shipping FR202 runtime are still being qualified and are not presented as a hard bound. Start a pilot at a conservative task period, then measure with the same devices, communications, logging, and environment expected in service.
Reading the results
Lower execution time means more headroom; it does not by itself make a complete controller more reliable. Judge control timing by the tail and the number of missed deadlines, not only by an average. See PLC Timing and Fault Behavior before planning a pilot.