Early Detection of Failure in Solid State Devices Using 4 Corner Targeted Testing

Abstract

The 4-Corner testing suite (4C) validates the drive's ability to perform under stress by cycling the drive through four extreme corners of thermal and voltage stress (4C- TVS) and periodic sudden power loss at temperature (4C-SPL), all while stressing the drive with workload. The test conditions such as voltage fluctuations, temperature above the product specification are the accelerating factors that simulates the real-world failures that the drive may encounter during product life cycle. The early phases of testing firmware maturity rely on bench validation tests to evaluate the stability of the firmware. These tests are targeted tests that validate the system as a whole unit and may not cover the temperature and voltage ranges that the drives are specified for including margins. Running the 4C-tests during the early phase of the project helps to uncover issues that may remain unnoticed until later phases of the project. Reliability SSD validation cycle involves testing the stability and performance of the drive or DUT at various stages of product development. There are 4 stages identified in line with the product maturity that are Pre-Engineering verification Testing (Pre-EVT), Engineering Verification Testing (EVT), Design Verification Testing (DVT) and Reliability Demonstration Test (RDT). Various tests are run at each stage of product development to identify issues, debug failures, and validate the fix. This process is repeated until the product is stable. The time between the Pre-EVT and DVT phase is about six months. The benefit of the 4C testing suite during the early firmware development phases such as Pre- Evt has seen an increased improvement in product stability. Early discovery allows for more comprehensive firmware and hardware solutions and better validation. •Reduces the debug time in later stages of the product development where issues are more complex to fix. •Lesser surprise failures in later stages of the product. •Better stability and confidence in the product. The target tests during the pre- EVT phase are bench validation test that test the drive feature by feature on limited drive population. The 4C tests on the other hand test the drive as a whole entity on larger volume of drives. The increased drive count and stressful environment increases the possibility of catching the failures as compared to bench validation tests. In one of the product evaluations, 4C was run during the pre-EVT phase. The run detected many failures such as power failures, bad block count increase, firmware issues etc., that were not caught by the bench validation tests. The issues uncovered during the run helped in improving the bench validation tests and increased overall quality of the product.