PROJECT TITLE :
Reimagining Heterogeneous Computing: A Functional Instruction-Set Architecture Computing Model
The relentless push in technology scaling driven by Moore's law has witnessed fantastic gains in the quantities of transistors accessible on chips. Laptop architects have exploited the additional transistors by incorporating many computing cores at intervals a single processor. Heterogeneous processing in explicit has become a helpful technique for handling ever-present power and memory restrictions. Yet, the scope and variety of current heterogeneous designs stay bounded by the amount of practical abstraction specified by conventional instruction-set architectures (ISAs). In this text, the authors demonstrate how the functional abstraction level determines the aptitude and variety of a processor's functional units and accelerators, thereby limiting its degree of heterogeneity. Combining current heterogeneous techniques with software abstraction ideas, the authors propose a replacement useful ISA (F-ISA), which raises the purposeful abstraction level of machine instructions. Using this model to enhance existing architectures makes accessible a wider scope and diversity of practical units and accelerators in order to use the ever-increasing transistor densities. Larger heterogeneity can offer advances in terms of object data mapping and execution, ensuing in potentially substantial latency, memory footprint, and power/performance gains.
Did you like this research project?
To get this research project Guidelines, Training and Code... Click Here