Abstract: Instruction and data registers of processors of a multiprocessing computing system are joined and forked to allow processing in multiple modes of operation. When joined, the registers of the processors each contain a same piece of information, hence generating single instruction and data streams. In contrast, when forked, the registers of the processors contain different pieces of information, thereby generating multiple instruction and data streams. Additionally, information may be stored into partitions of memory and fetched and broadcast by processors local to the particular memory sections thereby resulting in a faster cycle time.

Abstract: A job scheduler module for a distributed processing system is provided with capability to protect jobs from planned or unplanned subsystem downtimes. The approach includes determining a downtime for a subsystem of the distributed processing system, determining at least one of a start time and an end time of a job to be executed using the subsystem, determining whether the start time or the end time of the job is within the downtime, and if not, placing the job in an eligible job list. The scheduler then makes a decision whether to start the job using the eligible job list. A downtime protection flag can be associated with each job, thereby allowing a user to indicate whether the job is to receive subsystem downtime protection. If not, then the job is placed directly in the eligible job list for scheduler execution irrespective of subsystem downtime.