13 January 2023 in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
As the demand for data-intensive analytics has increased with the rapid advance in artificial intelligence (AI), various AI accelerators have been proposed. However, as AI-based solutions have been adapted to applications requiring accuracy and reliability, the reliability of them has become a critical issue. For this reason, self-test and self-recovery for AI accelerator (STRAIT) is proposed in this article. It facilitates self-test, self-diagnosis, and self-recovery by utilizing the structural and operational characteristics of systolic array in AI accelerator. The proposed self-test is progressed using scan chains composed of functional paths and can achieve a 100% test coverage (for both stuck-at and transition-delay faults) with a small number of test patterns and reduced test power. The proposed self-diagnosis is progressed with the proposed self-test in real time and allows accurate fault localization with fault type analysis. The proposed self-recovery is progressed using efficient pruning for faulty processing elements with weight allocation, and the reliability of AI accelerators can drastically increase with negligible performance degradation. However, STRAIT can be implemented with a small area overhead.
27 August 2021 in IEEE Access
Scan chain diagnosis is essential to solving yield-reduction problem caused by the miniaturization of manufacturing process. The accurate diagnosis of scan chain faults that frequently occur in the initial process is vital for rapidly improving yield. Moreover, the importance of scan chain diagnosis with a high resolution for the multiple faults is increasing because multiple faults occur in the early stages of the process, further increasing the cost of physical failure analysis. Although multiple faults can be diagnosed with existing methods, a high diagnostic resolution is difficult to achieve in the early stages of the process (where many faults occur) due to the rapid increase in the number of diagnosed fault candidates as the number of actual faults in the circuit increases. In this paper, a novel reconfigurable scan architecture that reconfigures the diagnosis paths and a test algorithm that uses this scan architecture are proposed to reduce the number of diagnosed fault candidates in the scan chain diagnosis with multiple circuit faults. Experimental results indicate that the proposed method achieves the higher diagnostic resolution for multiple faults than conventional methods. In addition, the proposed method reduces the routing overhead by scan partitioning.
3 June 2019 in IEEE Transactions on Very Large Scale Integration (VLSI) Systems
Low-power design is a key consideration in modern design. xor self-gating (data-driven self-gating) is used for power reduction in clock networks, which is one of the main factors of dynamic power consumption. When applying xor self-gating, dynamic power consumption is reduced, but the number of required test patterns on the testing side is inflated. In critical cases, more than three times the regular number of scan test patterns may be required for industrial designs, such as GPUs. In this brief, we propose a novel self-gating structure. Data- selectable self-gating (DSSG) is designed to use functional data and scan data selectively to eliminate the unnecessary clock toggling of flip-flops. With this structure, the self-gating function can be used in the scan test mode, as well as the function mode. When the self-gating logic is used during scan shift operations, the stuck-at faults in the self-gating logic can be tested with short test sequences; therefore, the rise in test costs can be mitigated. It is possible to test the stuck-at faults in self-gating logic using only four scan test patterns. The experimental results show that the average of the stuck-at test pattern increase ratio has been dropped from more than 90% to less than 8%. The low-power performance of the proposed method in the mission mode is the same as that of the conventional self-gating structures. When the DSSG method is used, the dynamic power of the shift operation which may increase excessively during the scan test can be reduced.
7 September 2020 in IEEE Access
Nanotechnology is an important technological alternative to overcome the limitations of complementary metal-oxide-semiconductor (CMOS) technology. Various circuit implementation methods based on nanotechnology have been proposed, and their most important characteristics are a high defect ratio and defect tolerance through circuit reconfiguration. CMOS-nanowire-MOLecular (CMOL) field-programmable gate array (FPGA) circuits are advanced logic circuit structures that combine the advantages of CMOS and nanotechnology. However, researches on defect diagnosis methods for the reconfiguration of CMOL FPGA circuits are barely conducted. In this paper, efficient circuit configuration methods for defect diagnosis of the CMOL FPGA circuits are proposed to address the problem. Also, diagnosis algorithms for both stuck-at open and stuck-at close defects are introduced. Experimental results show that with the proposed methods, diagnosis is possible for CMOL fabrics with up to 20% stuck-at open defects and 0.02% or more stuck-at close defects.