Good day to all,
From an architecture point of view is it fair to assume there can be
only one timestamp generator per system or there can be more, something like one per CPU cluster? Either way, is the current practice expected to continue or there are plans to move to different design standards?
Thanks, Mathieu