“Real-time” – A Misconstrued Term

July 29th, 2013 by

By Daron Underwood

Someone from our traditional markets probably read the title of this post and said, “I know what real-time means.”  That’s probably true in respect to the term’s use in real-time operating systems for embedded control.

But I’m not talking about real-time in the traditional sense here.  I’m referring to the use of the RTX Platform in markets, where “real-time” means something significantly different.

For instance, let’s take a look at what we call “Digital Media” market where the RTX Platform – Windows plus RTX running on multicore multiprocessor – continues to gain traction.

The real-time and deterministic behavior of the RTX real-time subsystem are essential in Digital Media critical-processing-execution environments.

What do I mean by the term “critical-processing-execution environment”?  Basically, RTX separates the processing of critical threads from the processing of normal Windows threads – both user mode and kernel mode. Windows and RTX are essentially subsystems symbiotically sharing the same resources in a mutually stable way. The platform allows for the system to be configured to minimize/eliminate, the influence of one subsystem on the other.

OK, so now that there is a basic understanding of how RTX can play a role in these non-traditional applications, let’s get back to the topic at hand. Real-time systems in the embedded sense, generally refer to a systems ability to respond to events in a bounded time frame.  For instance, if a switch is toggled at time “t”, the system is designed to perform some operation “o” within a given time window and be finished with that operation within a given time duration.  These events may or may not occur at known time intervals.  Regardless of expected or not, the system must respond in a deterministic way.

Returning to “Digital Media,” think thinking now about delivering streaming audio over a network. There is a need to provide data in a fashion that allows the stream to continue at a required rate.  RTX provides the environment which supports very low latency processing of this type of data as well as the ability to easily scale to more and more streams by utilizing multicore hardware.

In summary, even though the terminology might be different, the underlying requirement to process critical data requires a processing environment that can provide a guaranteed performance level to support many different execution models. The RTX Platform provides just such an environment, bolstered by the flexibility and scalability of multicore support.

Don’t get caught up in terminology, just understand that there is a platform that can support scenarios envisioned here…and many others.


Leave a Comment