Ali U. Irturk
Ph.D. at University of California, San Diego
M.A. and M.S. at University of California, Santa Barbara
B.A. and B.S. at Turkish Naval Academy
The choice of a computing platform plays an important role in the overall design of systems. Previously, designers decided between a
hardware or software implementation. The hardware implementation consisted of designing an Application Specific Integrated Circuit
(ASIC), which offers exceptional performance, but long time to market and high costs for all but the largest production chips.
The software route tended towards the use of Digital Signal Processors (DSPs) due to the ease of development and fast time to market.
However, they lack the performance for high throughput applications. Recently Field Programmable Gate Arrays (FPGAs) have become
prevalent for computationally intensive applications. FPGAs play a middle role between ASICs and DSPs, as they have the programmability
of software with performance approaching that of a custom hardware implementation. However, FPGAs require vast amounts of customization
throughout the design process and few tools exist which can aid the designer with the many system, architectural and logic design
choices. Designing a high level tool for fast prototyping application specific architectures is crucial.
My research work concentrates on
1) Design of a tool, GUSTO, for automatic generation and optimization of application specific architectures;
2) Acceleration of financial computations with FPGA architectures that are implemented using GUSTO;
3) Rapid architectural production of various wireless communication, signal processing and computer vision algorithms using GUSTO.