A TEMPERATURE AWARE Z-CURVE BASED ONLINE TASK PLACEMENT ALGORITHM FOR PARTIALLY RECONFIGURABLE FPGA s
Senoj Joseph, K. Baskaran · 2014
Partially runtime reconfigurable FPGA’s allow hardware tasks to be placed and removed dynamically at runtime. A fast and efficient algorithm for finding empty area is necessary for online placement algorithms. This paper deals with online scheduling and placement of tasks onto partially reconfigurable FPGAs in which CLB are labeled using Z-curve. The free space can be described easily using one dimensional run length based coding thereby making addition and deletion of task as simple as inserting entries into the run length list. Simulations indicate that the proposed methods produce better placement with 10% less task rejection when compared with related approaches Nowadays temperature distribution on the FPGA is a major issue. Hence the work is extended to develop a novel temperature aware model. This algorithm divides the entire surface of the FPGA into several clusters. Task are allocated to each cluster based on parameters like temperature, stress, fragmentation etc. The proposed model achieves lower temperature distribution with slight increase in rejection rate.