non-commercial and non-profit purposes only.

Heterogeneous computing environments use an assortment of high performance ma-chines with different architectures in an attempt to most efficiently execute the various tasks that are required by an application. While this environment is very well suited for tasks such as image understanding, heterogeneous processing has key limitations which including the lack of support for fine-grained parallelism, the high communica-tion overhead of moving tasks between machines, and the prohibitively high costs. The goal of this thesis is to propose a new computing paradigm, called micro-Heterogeneous computing or mHC, which incorporates PCI based processing elements (vector processors, digital signal processors, etc) into a general purpose machine. In this manner the benefits of heterogeneous computing on scientific applications can be achieved while avoiding some of the limitations. Overall performance is increased by exploiting fine-grained parallelism on the most efficient architecture available, while reducing the high communication overhead and costs of traditional heterogeneous environments. Furthermore, mHC based machines can be combined into a cluster