As simulation and training systems become more complex, vendors must rely on the ability of the target system to meet the processing needs of the application. The ever increasing complexity of today's training systems has exceeded the processing capabilities of many single CPU systems. As an alternative, more and more vendors are now considering multi-processor systems.

The Ada language is the logical choice as a software environment for developing these large scale applications. The Ada tasking mechanism can be extended to schedule and distribute tasks over multiple processors. This resulting parallel Ada runtime is capable of executing Ada tasks in parallel, while upholding the rules of the Ada language.

The decision to migrate to a parallel Ada environment is an important one involving many important factors. The intention of this study is to provide the applications developer with an insight into the specific features available in parallel Ada environments, and which features will be most useful throughout the life cycle of his application. With this information, the decision maker should be able to determine if a parallel Ada target environment is worth considering, and which types of parallel environments provide the individual features most essential to the success of his application.