Design of programmable memory BIST using stochastic sobol sequence pattern generator

The development of IC integration technologies leads to an extensive use of memories and buffers in different memory intensive applications. Therefore, probability of occurrence of fault in every single read and writes operation is increased in Memory BIST (MBIST). There were many sobol sequence testing approaches that were developed for efficient sobol sequence testing and diagnosis of fault. However, all algorithms are not strengthened enough to detect all possible faults that may be present due to fabrication errors or environmental disturbance. Keeping this in mind and taking the possibility of development of efficient algorithm a hybrid memory sobol sequence testing algorithm is presented. To overcome those drawbacks, Pipelining based MBIST designed to detect the all the types of memory faults by utilizing March-C sobol sequence testing algorithm. By introducing the Pipelining approach, majorly path delays are reducing. The proposed architecture designed and verified using Xilinx ISE environment under various sobol sequence testing methods with respect to the different category of memories. The simulation and synthesis results shows that the proposed method shows the enhanced performance with the hardware resource utilization and delay consumption compared to the conventional approaches.