Mr. van der Wal, PLCopen normally deals with software environments in automation technology. How is it that the association is now setting out to define and implement a benchmark for CPU architectures?

Eelco van der Wal: Yes, that's not really our playground, you're right. But it seemed that there was no other interest group that could take on this matter. So we became active ourselves.

What are the most important findings of the benchmark?

Eelco van der Wal: We are not interested in generating a processor comparison as such. Rather, we want to create a benchmark tool. We are developing different test scripts - and not just for the real-time issue, but also for everything else around it: memory access, caching, PCI Express access, communication via OPC UA and power consumption.

Probably the most interesting finding from all this work is that when creating the benchmark, it became clear that both the CPU suppliers and the controller manufacturers benefit from finally speaking the same language and thus gaining a much better understanding of each other.

Who can ultimately benefit from the benchmark?

Eelco van der Wal: Both the CPU suppliers benefit because they can now adapt their CPU implementations to the requirements of the controller suppliers. And the controller manufacturers benefit from the CPUs, which are better tailored to their needs thanks to the benchmark.

Will there be follow-up projects?

Eelco van der Wal: We are currently in the first phase of the project: building mutual understanding and specifying the initial test requirements, including the first test scripts. In follow-up projects, we want to expand these tests, also to support the runtime environments for software from 3rd party providers.

Who can use the benchmark tool? Or in other words: for whom is it available?

Eelco van der Wal: The tool is currently only available to members of the working group. However, we would like to make it available to all PLCopen members in the near future.

We have not yet decided how openly we will ultimately release the tool and whether the test scripts will be available to the general public in Github, for example.
What we want to avoid in any case is that the tool is seen as a generic benchmark for positioning controllers from different manufacturers against each other.