The current advances in computer technology has resulted in manufacturing inexpensive, high-capacity, high-computationally-feasible systems. A cluster of these systems is capable of outpacing some of the high-performance supercomputers at a fraction of cost. Face recognition, belonging to the family of computationally intensive applications, requires high-performance systems to perform its calculations. To solve the calculations in near real-time would require expensive high-performance supercomputers or clustering inexpensive systems.
Therefore a scalable distributed multi-component framework system called Apollo is proposed. This system would augment the computational intensity requirments of face recognition matching on a large scale. Furthermore Apollo was designed with reliability, availability, and serviceability (RAS) in mind. The reliability lies in its dynamicity and adaptability traits in changing network environment. The off-site storage and redundancy fulfill the availability profile and the serviceability lies in its scalable feature and option in replacing (or upgrading) components.