To watch the Flash version, you need Flash 8 or later, and Javascript must be enabled in your browser.
[ MUSIC ]
Hi, I'm Fred Clewis with the IBM Linux Integration Center. As part of the IBM Business Critical Linux Series, this segment focuses on Real-Time Linux and Java, a component of the Next Generation Infrastructure. We'll look at the basics of real-time computing, and how Linux and Java have greatly lowered the cost for enterprises that use or want to start using real-time computing.
Real-time computing can now spread beyond traditional industries into industries that simply want a competitive edge. Let's take a closer look.
Real-Time Linux includes the regular benefits of Linux, with the addition of specific features designed for real-time computing. Real-time computing -- also known as deterministic computing -- is required for computations that must complete in a predictable time period.
A real-time system is one in which the correctness of a computation not only depends on the logical correctness of the computation but also on the time at which the result is produced.
Real-time systems are required for many industries, including financial services, manufacturing and military. Most modern computer operating systems use a process known as preemptive multi-tasking. The Central Processing Unit, or CPU, works on many unrelated tasks, spending a small amount of time on each one to give the illusion of simultaneous execution.
This time period is known as a time slice, and a program thread will usually require many time slices to complete. Time slices are defined by periodic interrupt events that are sent to the CPU.
The CPU cannot immediately respond to an interrupt event; some housekeeping must be done to stop the current task, determine which tasks to run next and prepare for the new task. The time between the arrival of the interrupt event and the start of the new task is called the response time.
For real-time computing, the response time must always be deterministic and operate within a known worse case time, regardless of system load. Many non real-time systems can be fast enough some or even most of the time, but cannot guarantee to be fast enough every time.
Real-time solutions have been around for years; however, they were custom built and required specialized hardware, specialized operating systems and specialized skills. Applications developed in these environments were difficult to reuse or report to other platforms. Previous real-time solutions have been both proprietary and costly.
IBM recognized the need for a real-time solution using low cost commercially available open, scalable and flexible technologies, including X86-based hardware, standard Linux distributions with real-time capabilities, and the Java programming language.
For this to happen, real-time issues with Linux and Java would have to be overcome. Linux had unpredictable scheduling, a low timer resolution and a non-preemptable kernel. Java had issues with thread scheduling, memory garbage collection and class compiling and loading.
IBM worked with the Open Source community to overcome these issues, and today offers a real-time solution that is fast, predictable and easy to manage. And, it is available at significantly lower cost than proprietary solutions.
The solution starts with select IBM hardware, with firmware enhancements to optimize real-time workloads. The operating system is a commercially available Real-Time Linux, which has the timing, interrupt and scheduling changes required by real time.
Java is provided by IBM's WebSphere Real Time, based on Java 2 Standard Edition, and in compliance with the real-time specification for Java. This Java version has memory management and compilation strategies required for real time.
While traditional real-time industries like finance and the military are seeing the initial benefits of Real-Time Linux and Java, there are numerous advantages for other industries, from government to healthcare to manufacturing.
Real-time processing provides predictability to applications such as real-time product simulations, language translations and audio/video streaming. For many industries seeking a competitive edge, the time is now to move to real-time processing. To learn more about how Real-Time Linux and Real-Time Java can benefit your organization, visit this site.
[ MUSIC ] [END OF SEGMENT]
