Spin-transfer torque magnetic RAM (STTMRAM) is a promising candidate to replace SRAM due to its beneficial features of non-volatility, higher density, and zero leakage power. To further increase the data density, multi-level cell (MLC) STT-MRAM, which can store two bits per cell has been proposed. However, it has to conquer the two-step transition (TT) threats for some state transitions because writing the hard domain of a cell also writes the soft domain to the same value. The main drawbacks of TT are higher power consumption and a performance penalty. To address this issue, this paper proposes a novel two-step transition zeroization (TSTZ) technique that can completely eliminate two-step transitions. TSTZ is basically a ternary encoding technique where each MLC cell stores a ternary digit. The ternary digits 0,1, and 2 are encoded into cell states 00, 01/10, and 11, respectively. Since only three states of an MLC cell are used at a time, TSTZ eliminates TTs and significantly reduces the read error rate. Corresponding hardware architectures for encoding and decoding have also been proposed. The space overhead incurred is about 33%, the lowest among existing techniques. Power consumption and lifetime can be improved by 12% and 22.1% respectively. Moreover, the read error rate can be reduced from the original 8.24% to 0.75%.

Deep neural networks (DNNs) are being widely used in smart appliances, face recognition and autonomous driving. The trained weight data are usually stored in flash memory which suffers from reliability and endurance issues. Owing to the inherent error tolerability for DNN applications, address remapping techniques are proposed for protecting weight data stored in flash memory. Bit significances are first analyzed and then a weight transposer is proposed for remapping significant weight bits to fault-free or much reliable flash cells. A bipartite graph model is developed for modeling address remapping. The corresponding hardware architectures for address remapping are also proposed. We use the deep learning framework pytorch for evaluating inference accuracy for different DNN models. Experimental results show that based on 0.01 % injected BER in the weight data, the accuracy losses of widely used DNN models are less than 1 % with negligible hardware overhead.

High power consumption in LSI testing may cause excessive IR-drop. When IR-drop becomes excessive, it causes excessive delay, resulting in test malfunction (over-testing). Excessive IR-drop does not occur in the entire area of a circuit, but in certain areas where a large number of switching activities occur (such areas are called hotspots in this work). In order to avoid test malfunction, it is important to develop a method to reduce or control IR-drop in the hotspots. Locating hotspots is a necessary technique to reduce or control IR-drop effectively and efficiently. In this work, we propose a method to locate hotspots in a logic circuit by switching probability calculation. Experimental results for IWLS2005 OpenCores circuits demonstrate the proposed method can support to locate hotspots.

In this paper, we propose a practical error detection method for combinatorial circuits using three-site implications which satisfies safety integrity level ASIL-B of ISO 26262 standard. The proposed method finds implications which are invariant relationships among internal signals in a logic circuit for any input pattern, and adds checkers to confirm that the implications are satisfied during functional operation. While the checkers playa role of CED (Concurrent Error Detection), higher error detection coverage can be achieved by using three-site implications than only with two-site implications as is the case in previous works. However, as circuit size grows it becomes difficult to find out effective three-site implications in reasonable processing time due to a large number of candidate implications. In the proposed method, along with various speed up techniques utilizing commercial EDA tools, newly developed indexes to narrow down the search space in three-site implication extraction with minimum loss of error detection coverage can achieve reasonable processing time, error detection coverage and area overhead. As preliminary experiments with resynthesized ISCAS85 benchmark circuits, we confirmed that the error detection coverage using all three-site implications under the limit of 50% area overhead is 19.5% higher than only with two-site implications, and the checkers using three-site implications extracted by the proposed indexes can achieve 96% of the error detection coverage derived using all three-site implications. We also applied the proposed method to an industrial CPU core, tinyMicon MatisseCORE™, as complementary error detection of software-based diagnosis. Through this experiment, we confirmed that 90% diagnostic coverage, which is equivalent to ASIL-B requirement in ISO 26262 standard, can be achieved by the checkers added by the proposed method with 15.41 % area overhead. This result implies the proposed method can be applied to actual commercial products.

Built-in self-repair (BISR) techniques has been considered as the most cost-effective solution for enhancing yield and reliability of NAND flash memory. Owing to the inherent architecture of NAND flash memory, conventional BISR techniques use spare columns and NAND blocks as the basic replacement elements. These techniques can be categorized as the coarse-grained BISR techniques (CGBISR). It is evident that the efficiency of spare usage is very low. To cure this dilemma, fine-grained BISR (FGBISR) techniques are proposed in this paper. We first exploit the fault behaviors at the circuit level and derive novel and concise repairable fault types (RFTs) for the widely used flash memory fault models. The proposed RFTs include bit-, page-, column-, and NAND block-repairable faults. Therefore, FGBISR can conduct repairing at the fine-grained levels for improving repair efficiency. We also provide efficient redundancy analysis algorithms suitable for VLSI implementation based on the RFTs. The corresponding FGBISR architectures and repair flow are also proposed. A simulator was developed for evaluating repair rate, yield, reliability, and hardware overhead. Experimental results show that repair rate, yield, and reliability can be raised significantly with negligible hardware overhead.

Deep neural networks (DNNs) are being widely used in smart appliances, face recognition and autonomous driving. The trained weight data are usually stored in flash memory which suffers from reliability and endurance issues. Owing to the inherent error tolerability for DNN applications, adaptive address remapping techniques are proposed for protecting weight data stored in flash memory. Bit significances are first analyzed to determine the priority of weight bits which should be protected. Thereafter, a novel weight transposer and an address remapper are proposed for remapping significant weight bits to fault-free or much reliable flash cells. A bipartite graph model is developed for modeling address remapping and evaluating error score. The corresponding hardware architectures for address remapping are also proposed. We use the deep learning framework pytorch for evaluating inference accuracy for different DNN models. Experimental results show that based on 0.01 % injected BER in the weight data, the accuracy losses of widely used DNN models are less than 1 % with negligible hardware overhead.

Performance degradation caused by high IR-drop in normal functional mode of LSI can be avoided by improving the power supply network in the layout design phase. However, while IR-drop increases much more in test mode than in normal functional mode, excessive IR-drop in test mode is not appropriately considered in the layout design phase. Excessive IR-drop in test mode causes over-testing, which wrongly determines a fault free LSI in normal functional mode to be faulty. In this work, we propose a method for analyzing high IR-drop areas (hotspot distribution), which is necessary to effectively and efficiently reduce excessive IR-drop.

Phase change memory (PCM) is considered as the most promising alternative of DRAM. However, it has the inevitable endurance problem of the storage cells. The limited endurance and other permanent faults cause serious reliability and yield challenges. Conventional techniques like hard repair schemes and error correction codes (ECC) are usually used to overcome these dilemmas. However, since soft errors are not a main threat for PCM, equipping ECC for each data word will waste a lot of memory space for storing the check bits. Therefore, progressive ECC techniques are proposed to solve this drawback in this paper. The main idea is to equip ECC for data words when their first faulty bits are detected. That is, only the fault detection code is equipped for data words such that the original code rate is high. A separated ECC DRAM is used for storing the check bits. Two types of progressive ECC techniques¾the local progressive ECC (LPE) technique and the global progressive ECC (GPE) technique are presented. The proposed techniques are also easy to be integrated with the conventional BISR (Built-in Self-repair) architectures. According to experimental results, the degradation of repair rate and reliability are almost negligible. However, the hardware overhead is at least 70% lower than the original ECC technique.

High scan shift power often leads to excessive heat as well as shift timing failures. Partial shift (shifting a subset of scan chains at a time) is a widely adopted approach for avoiding excessive heat by reducing global switching activity, we show for the first time that it may actually cause excessive IR-drop on some clock buffers and worsen shift clock skews, thus increasing the risk of shift timing failures. This paper addresses this problem with an innovative method, namely Clock-Skew-Aware Scan Chain Grouping (CSA-SCG). CSA-SCG properly groups scan chains to be shifted simultaneously so as to reduce the imbalance of switching activity around the clock paths for neighboring scan flip-flops in scan chains. Experiments on large ITC'99 benchmark circuits demonstrate the effectiveness of CSA-SCG for reducing scan shift clock skews to lower the risk of shift timing failures in partial shift.

The yield and reliability issues are important challenges for the emerging phase change memory (PCM). Hard repair techniques based on fault replacement and error correction codes are usually used to cure these dilemmas. However, the probability of occurring permanent faults is low and soft errors are not a main threat for PCM, equipping ECC for each data word will waste a lot of storage space. Therefore, an adaptive ECC technique is proposed to solve this drawback. The main idea is to equip ECC for memory words when they are detected faulty. A separated ECC DRAM is used for storing the check bits. According to experimental results, the degradation of repair rate is almost negligible. However, the hardware overhead is at least 70% lower than the original ECC technique.

Loading and unloading test patterns during scan testing causes many scan flip-flops to trigger simultaneously. This instantaneous switching activity during shift in turn may cause excessive IR-drop that can disrupt the states of some scan flip-flops and corrupt test stimuli or responses. A common design technique to even out these instantaneous power surges is to design multiple scan chains and shift only a group of the scan chains at a same time. This paper introduces a novel algorithm to optimally group scan chains so as to minimize the probability of test data corruption caused by excessive instantaneous IR-drop on scan flip-flops. The experiments show optimal results on all large ITC'99 benchmark circuits.

Excessive IR-drop during scan shift can cause localized IR-drop around clock buffers and introduce dynamic clock skew. Excessive clock skew at neighboring scan flip-flops results in hold or setup timing violations corrupting test stimuli or test responses during shifting. We introduce a new method to assess the risk of such test data corruption at each scan cycle and flip-flop. The most likely cases of test data corruption are mitigated in a non-intrusive way by selective test data manipulation and masking of affected responses. Evaluation results show the computational feasibility of our method for large benchmark circuits, and demonstrate that a few targeted pattern changes provide large potential gains in shift safety and test time with negligible cost in fault coverage.

Launch-Switching-Activity (LSA) is a serious problem during at-speed testing of integrated circuits, since localized LSA may lead to severe IR-drop and thus failures. The excessive LSA is conventionally mitigated by reducing the switching activity through special low-power test generation techniques, typically resulting in severe test pattern inflation and high test costs. This work introduces a novel concept of Low-Capture-Power Test Points (LCP-TPs), which are inserted to reduce switching activity in critical High-Capture-Power (HCP) regions. LCP-TPs also help in retaining high test compaction capability. An optimization- SAT based procedure is proposed to compute a small set of optimal LCP-TP locations for compact at-speed test sets with effective capture power reduction. Experimental results clearly demonstrate the advantages of LCP-TP insertion.

Detailed knowledge of a circuit's timing is essential for performance optimization, timing closure, and generation of test patterns to detect small-delay defects. When an input transition is applied to the circuit's inputs, the resulting delay is not only determined by the propagation path, but also influenced by the power-supply noise. We introduce a path-sensitization procedure which precisely controls the switching activity in the circuit region surrounding the path. The procedure can maximize or minimize switching activity, or set it to a user-specified value. We study the accuracy-vs.-efficiency trade-offs for a hierarchy of timing models, from coarse zero-delay assumption to a waveform-accurate approach with sub-cycle resolution. For the first time, we present a MaxSAT formulation which guarantees maximization or minimization of switching activity, stemming from transitions and from glitches, simultaneously with path sensitization. We validate the quality of the generated test patterns using a mixed-mode IR-drop-aware timing simulator.

With more and more sophisticated low-power design techniques being applied to modern LSI chips for aggressive functional power reduction, the risk of fault-free chips falsely failing production test grows due to excessively high test power compared with functional power. Existing low-power ATPG methods, however, suffer from severe test data inflation and often use unfocused global test power reduction. This paper proposes a novel optimization-SAT-based at-speed scan test generation method that is explicitly targeted at eliminating high-capture-power test vectors in a pre-generated compact test set. This method employs layout information in reducing capture switching activity in a focused regional manner. Experiments demonstrate that the proposed method can effectively eliminate a large number of high-capture-power test vectors with neither test data inflation nor fault coverage loss.

IR-drop induced by launch switching activity (LSA) in capture mode during at-speed scan testing increases delay along not only logic paths (LPs) but also clock paths (Cps). Excessive extra delay along LPs compromises test yields due to false capture failures, while excessive extra delay along CPs compromises test quality due to test clock stretch. This paper is the first to mitigate the impact of LSA on both LPs and CPs with a novel LCPA (Logic/Clock Path-Aware) at-speed scan test generation scheme, featuring (1) a new metric for assessing the risk of false capture failures based on the amount of LSA around both LPs and CPs, (2) a procedure for avoiding false capture failures by reducing LSA around LPs or masking uncertain test responses, and (3) a procedure for reducing test clock stretch by reducing LSA around CPs. Experimental results demonstrate the effectiveness of the LCPA scheme in improving test yields and test quality.

Power-related problems in at-speed scan testing have become more and more serious, since excessive IR-drop caused by excessive power consumption results in overtesting. There are two important factors in low-power testing: one is power estimation, the other is power reduction. Several estimation methods have been proposed based on the analysis of switching activity characteristics. In order to estimate the impact of IR-drop, it is more important to consider the area containing many cells which consume excessive power than to consider the total number of switching activity in a circuit. In this paper, we propose a novel method for identifying areas where excessive IR-drop likely occurs without using test vectors. Visualized experimental results for IWLS 2005 benchmark circuits demonstrate that the proposed method can effectively identify areas containing many cells which consume higher power than others. Such areas identified can be used in low-power test generation so as to achieve effective and efficient results.

This paper proposes a novel soft-error tolerant TCAM using partial don't-care keys (X-keys), namely TX, which significantly enhances the tolerance of the TCAM against soft errors. Experimental results show that the soft-error tolerance of the TX outperforms existing schemes. Moreover, the overhead of the TX is very small.

Ongoing research to shrink feature sizes of LSI circuits leads to an always increasing number of logic gates in a circuit. In general, the complexity of test generation depends on the size of a circuit. Furthermore, modern test generation methods have to consider power reduction in addition to fault detection, since excessive power caused by testing may result in over testing. In this work, we propose a method to reduce the computation time of low-power test generation. The proposed method specifies gates which will cause power issues, consequently reducing the search space for X-filling technique. The reduction of search space for Xfilling also further minimizes the amount of switching activity. Experimental results for circuits of Open Cores provided by IWLS2005 benchmarks show that the proposed method achieves both a reduced computation time and at the same time increased power reduction compared to previous methods.

Low-power test technology has been investigated deeply to achieve an accurate and efficient testing. Although many sophisticated methods are proposed for scan-test, there are not so many for logic BIST because of its uncontrollable randomness. However, logic BIST currently becomes vital for system debug or field test. This paper proposes a novel low power BIST technology that reduces shift-power by eliminating the specified high-frequency parts of vectors and also reduces capture power. The authors show that the proposed technology not only reduces test power but also keeps test coverage with little loss.

High test power in logic BIST is a serious problem not only for production test, but also for board test, system debug or field test. Many low power BIST approaches that focus on scan-shift power or capture power have been proposed. However, it is known that a half of scan-shift power is compensated by test responses, which is difficult to control in those approaches. This paper proposes a novel approach that directly reduces scan-out power by modifying some flip-flops' values in scan chains at the last capture. Experimental results show that the proposed method reduces scan-out power up to 30% with little loss of test coverage.

This paper proposes a novel scheme to manage capture power in a pinpoint manner for achieving guaranteed capture power safety, improved small-delay test capability, and minimal test cost impact in at-speed scan test generation. First, switching activity around each long path sensitized by a test vector is checked to characterize it as hot (with excessively-high switching activity), warm (with normal/functional switching activity), or cold (with excessively-low switching activity). Then, X-restoration/X-filling-based rescue is conducted on the test vector to reduce switching activity around hot paths. If the rescue is insufficient to turn a hot path into a warm path, mask is then conducted on expected test response data to instruct the tester to ignore the potentially-false test response value from the hot path, thus achieving guaranteed capture power safety. Finally, X-restoration/X-filling-based warm-up is conducted on the test vector to increase switching activity around cold paths for improving their small-delay test capability. This novel approach of pinpoint capture power management has significant advantages over the conventionalapproachofglobalcapturepower management, as demonstrated by evaluation results on large ITC'99 benchmark circuits and detailed path delay analysis.

Excessive capture power in at-speed scan testing may cause yield loss due to timing failures. Although reducing the number of clock domains that capture test responses simultaneously is a practical and scalable solution for reducing capture power, no available capture-safety checking metric can assess its effect in an accurate-enough manner, especially when multiple clock domains capture test responses in a short period of time. This paper proposes a novel CLEAR (CLock-Edge-Arrival-Relation-based) capture-safety checking method that, for the first time, takes clock edge arrival times for different clock domains into consideration. The accuracy and usefulness of the proposed method have been clearly demonstrated by simulation-based evaluation with the largest ITC'99 benchmark circuit as well as real-chip-based evaluation with an industrial chip embedded with on-chip delay measurement circuitry.

Launch-off-Capture (LOC) and Launch-off-Shift (LOS) are the two main test schemes for at-speed scan delay testing. In the literature, it has been shown that LOS has higher performance than LOC in terms of fault coverage and test length, but higher peak power consumption during the launch-to-capture cycle. Power reduction seems to be the key to really exploit LOS test scheme. However, it has been proven that reducing too much test power can lead to test escape due to under-test. In this context, this study proposes a smart X-filling framework able to adapt peak power consumption during the launch-to-capture cycle according to the functional power, i.e. the power consumption of the circuit in functional mode. Here, the main goal is to obtain a final test set with peak power consumption as close as possible to the functional power. Experimental results, carried out on the well-known ITC'99 benchmarks, prove the feasibility of the proposed approach.

Unknown (X) values, originating from memories, clock domain boundaries or A/D interfaces, may compromise test signatures and fault coverage. Classical logic and fault simulation algorithms are pessimistic w.r.t. the propagation of X values in the circuit. This work proposes efficient hybrid logic and stuck-at fault simulation algorithms which combine heuristics and local BDDs to increase simulation accuracy. Experimental results on benchmark and large industrial circuits show significantly increased fault coverage and low runtime. The achieved simulation precision is quantified for the first time.

It has become necessary to reduce power during LSI testing. Particularly, during at-speed testing, excessive power consumed during the Launch-To-Capture (LTC) cycle causes serious issues that may lead to the overkill of defect-free logic ICs. Many successful test generation approaches to reduce IR-drop and/or power supply noise during LTC for the launch-off capture (LOC) scheme have previously been proposed, and several of X-filling techniques have proven especially effective. With X-filling in the launch-off shift (LOS) scheme, however, adjacent-fill (which was originally proposed for shift-in power reduction) is used frequently. In this work, we propose a novel X-filling technique for the LOS scheme, called Adjacent-Probability-based X-Filling (AP-fill), which can reduce more LTC power than adjacent-fill. We incorporate AP-fill into a post-ATPG test modification flow consisting of test relaxation and X-filling in order to avoid the fault coverage loss and the test vector count inflation. Experimental results for larger ITC'99 circuits show that the proposed AP-fill technique can achieve a higher power reduction ratio than 0-fill, 1-fill, and adjacent-fill.

Excessive capture power in at-speed scan testing may cause timing failures, resulting in test-induced yield loss. This has made capture-safety checking mandatory for test vectors. This paper presents a novel metric, called the TTR (Transition-Time-Relation-based) metric, which takes transition time relations into consideration in capture-safety checking. Capture-safety checking with the TTR metric greatly improves the accuracy of test vector sign-off and low-capture-power test generation.

High power consumption in scan testing can cause undue yield loss which has increasingly become a serious problem for deep-submicron VLSI circuits. Growing evidence attributes this problem to shift timing failures, which are primarily caused by excessive switching activity in the proximities of clock paths that tends to introduce severe clock skew due to IR-drop-induced delay increase. This paper is the first of its kind to address this critical issue with a novel layout-aware scheme based on scan segmentation design, called LCTI-SS (Low-Clock-Tree-Impact Scan Segmentation). An optimal combination of scan segments is identified for simultaneous clocking so that the switching activity in the proximities of clock trees is reduced while maintaining the average power reduction effect on conventional scan segmentation. Experimental results on benchmark and industrial circuits have demonstrated the advantage of the LCTI-SS scheme.

At-speed scan testing may suffer from severe yield loss due to the launch safety problem, where test responses are invalidated by excessive launch switching activity (LSA) caused by test stimulus launching in the at-speed test cycle. However, previous low-power test generation techniques can only reduce LSA to some extent but cannot guarantee launch safety. This paper proposes a novel & practical power-aware test generation flow, featuring guaranteed launch safety. The basic idea is to enhance ATPG with a unique two-phase (rescue & mask) scheme by targeting at the real cause of the launch safety problem, i.e., the excessive LSA in the neighboring areas (namely impact areas) around long paths sensitized by a test vector. The rescue phase is to reduce excessive LSA in impact areas in a focused manner, and the mask phase is to exclude from use in fault detection the uncertain test response at the endpoint of any long sensitized path that still has excessive LSA in its impact area even after the rescue phase is executed. This scheme is the first of its kind for achieving guaranteed launch safety with minimal impact on test quality and test costs, which is the ultimate goal of power-aware at-speed scan test generation.

NBTI, which is one of well-known aging phenomena, brings delay degradation in deep submicron VLSIs. In order to detect NBTI-induced delay faults, we need to estimate delay degradation and apply delay test for the circuit in the field. This paper discusses on estimation of NBTI-Induced delay degradation. We first analyze the effect of the delay degradation, and then give a procedure of path selection in which long paths after the delay degradation are selected for the delay test in the filed. Experimental results show that estimation of delay degradation significantly affects path selection, and accurate estimation is important for the test.

This study investigates the reasons why test power reduction through X-filling techniques works well for cycle-average power reduction but is not so efficient concerning instantaneous peak power reduction.

Reducing excessive launch switching activity (LSA) is now mandatory in at-speed scan testing for avoiding test-induced yield loss, and test set modification is preferable for this purpose. However, previous low-LSA test set modification methods may be ineffective since they are not targeted at reducing launch switching activity in the areas around long sensitized paths, which are spatially and temporally critical for test-induced yield loss. This paper proposes a novel CAT (Critical-Area-Targeted) low-LSA test modification scheme, which uses long sensitized paths to guide launch-safety checking, test relaxation, and X-filling. As a result, launch switching activity is reduced in a pinpoint manner, which is more effective for avoiding test-induced yield loss. Experimental results on industrial circuits demonstrate the advantage of the CAT scheme for reducing launch switching activity in at-speed scan testing.

Reducing IR-drop in the test cycle during at-speed scan testing has become mandatory for avoiding test-induced yield loss. An efficient approach for this purpose is post-ATPG test modification based on X-identification and X-filling since it causes no circuit/clock design change and no test vector count inflation. However, applying this approach to test compression has been considered challenging due to the limited availability of X-bits. This paper solves this serious problem by proposing a novel and practical CA (Compression-Aware) test modification scheme for reducing IR-drop in the widely-used broadcast-scan based test compression environment. This unique scheme features (1) CA circuit remodeling for minimizing the effort of applying test modification to broadcast-scan-based test compression, (2) CA X-identification for increasing X-bits for risky test vectors, and (3) CA X-filling for effectively using limited X-bits in reducing IR-drop. As a result, the CA test modification scheme can achieve significant IR-drop reduction even when a test cube only has a small number of X-bits. This advantage is clearly demonstrated by experimental results on three compression configurations created from an industrial circuit.

Power-aware X-filling is a preferable approach to avoiding IR-drop-induced yield loss in at-speed scan testing. However, the quality of previous X-filling methods for reducing launch switching activity may be unsatisfactory, due to low effect (insufficient and global-only reduction) and/or low scalability (long CPU time). This paper addresses this quality problem with a novel, GA (Genetic Algorithm) based X-filling method, called GA-fill. Its goals are (1) to achieve both effectiveness and scalability in a more balanced manner, and (2) to make the reduction effect of launch switching activity more concentrated on critical areas that have higher impact on IR-drop-induced yield loss.Evaluation experiments are being conducted on benchmark and industrial circuits, and initial results have demonstrated the usefulness of GA-fill.

Yield loss caused by excessive power supply noise has become a serious problem in at-speed scan testing. AlthoughX-filling techniques are available to reduce the launch cycle switching activity, their performance may not be satisfactory in the linear-decompressor-based test compression environment. This paper solves this problem by proposing a novel integrated automatic test pattern generation scheme that efficiently and effectively performs compressible low-capture-powerX-filling. Related theoretical principles are established, based on which the problem size is substantially reduced. The proposed scheme is validated by benchmark circuits, as well as an industry design in the embedded deterministic test environment.

At-speed scan testing is susceptible to yield loss risk due to power supply noise caused by excessive launch switching activity. This paper proposes a novel two-stage scheme, namely CTX (Clock-Gating-Based Test Relaxation and X-Filling), for reducing switching activity when test stimulus is launched. Test relaxation and X-filling are conducted (1) to make as many FFs inactive as possible by disabling corresponding clock-control signals of clock-gating circuitry in Stage-1 (Clock-Disabling), and (2) to make as many remaining active FFs as possible to have equal input and output values in Stage-2 (FF-Silencing). CTX effectively reduces launch switching activity, thus yield loss risk, even with a small number of donpsilat care (X) bits as in test compression, without any impact on test data volume, fault coverage, performance, and circuit design.

Capture-safety, defined as the avoidance of any timing error due to unduly high launch switching activity in capture mode during at-speed scan testing, is critical for avoiding test- induced yield loss. Although point techniques are available for reducing capture IR-drop, there is a lack of complete capture-safe test generation flows. The paper addresses this problem by proposing a novel and practical capture-safe test generation scheme, featuring (1) reliable capture-safety checking and (2) effective capture-safety improvement by combining X-bit identification & X-filling with low launch- switching-activity test generation. This scheme is compatible with existing ATPG flows, and achieves capture-safety with no changes in the circuit-under-test or the clocking scheme.

Test data modification based on test relaxation and X-filling is the preferable approach for reducing excessive IR-drop in at-speed scan testing to avoid test-induced yield loss. However, none of the existing test relaxation methods can control the distribution of identified don’t care bits (X-bits), thus adversely affecting the effectiveness of IR-drop reduction. In this paper, we propose a novel test relaxation method, called Distribution-Controlling X-Identification (DC-XID), which controls the distribution of X-bits identified from a set of fully-specified test vectors for the purpose of effectively reducing IR-drop. Experimental results on large industrial circuits demonstrate the effectiveness and practicality of the proposed method in reducing IR-drop, without any impact on fault coverage, test data volume, or test circuit size.

Yield loss caused by excessive power supply noise has become a serious problem in at-speed scan testing. Although X-filling techniques are available to reduce the launch cycle switching activity, their performance may not be satisfactory in the linear-decompressor-based test compression environment. This work is the first to solve this problem by proposing a novel integrated ATPG scheme that efficiently and effectively performs compressible X-filling. Related theoretical principles are established, based on which the problem size is substantially reduced. The proposed scheme is validated by large benchmark circuits as well as an industry design in the embedded deterministic test (EDT) environment.

IR-drop-induced malfunction is mostly caused by timing violations on activated critical paths during the capture cycle of at-speed scan testing. A critical-path-aware X-filling method is proposed for reducing IR-drop, especially on gates that are close to activated critical paths, thus effectively preventing test-induced yield loss.

High-quality at-speed scan testing, characterized by high small-delay-defect detecting capability, is indispensable to achieve high delay test quality for DSM circuits. However, such testing is susceptible to yield loss due to excessive power supply noise caused by high launch-induced switching activity. This paper addresses this serious problem with a novel and practical post-ATPG X-filling scheme, featuring (1) a test relaxation method, called path keeping X-identification, that finds don't-care bits from a fully-specified transition delay test set while preserving its delay test quality by keeping the longest paths originally sensitized for fault detection, and (2) an X-filling method, called justification-probability-based fill (JP-fill), that is both effective and scalable for reducing launch-induced switching activity. This scheme can be easily implemented into any ATPG flow to effectively reduce power supply noise, without any impact on delay test quality, test data volume, area overhead, and circuit timing.

X-filling is preferred for low-capture-power scan test generation, since it reduces IR-drop-induced yield loss without the need of any circuit modification. However, the effectiveness of previous X-filling methods suffers from lack of guidance in selecting targets and values for X-filling. This paper addresses this problem with a highly-guided X-filling method based on two novel concepts: (1) X-score for X-filling target selection and (2) probabilistic weighted capture transition count for Y-filling value selection. Experimental results show the superiority of the new X-filling method for capture power reduction.

High power dissipation can occur when the response to a test vector is captured by flip-flops in scan testing, resulting in excessive JR drop, which may cause significant capture-induced yield loss in the DSM era. This paper addresses this serious problem with a novel test generation method, featuring a unique algorithm that deterministically generates test cubes not only for fault detection but also for capture power reduction. Compared with previous methods that passively conduct X-filling for unspecified bits in test cubes generated only for fault detection, the new method achieves more capture power reduction with less test set inflation. Experimental results show its effectiveness.

Recently design for manufacturability (DFM) has been required to achieve higher process yield. Information obtained from silicon by testing and/or fault analysis is sometimes fed back for redesign of VLSI circuits. In this paper we propose a method to maximize defect coverage of a test set initially generated for stuck-at faults in a full scan sequential circuit by using feed back information from fault analysis. If a test set for more complex faults than stuck-at faults is generated, higher defect coverage would be obtained. Such a test set, however, would have a large number of test vectors, and hence the test costs would go up. The proposed method improves defect coverage of the test set by not adding new test vectors but modifying test vectors with the information obtained from fault analysis. Therefore there are no negative impacts on test data volume and test application time. The initial fault coverage for stuck-at faults of the test set is guaranteed with modified test vectors. In this paper we focus on detecting as many as possible non-feedback AND/OR-type bridging faults. Experimental results show that the proposed method significantly decreases the number of non-feedback AND/OR-type bridging faults undetected by a test set generated for stuck-at faults.

In this paper, we propose a method of test compression for multiple scan designs. Instead of the conventional serial scan chains, the proposed method constructs scan trees in which scan flip-flops are placed and routed in a tree structure. Inputs of the scan trees drive several scan trees of different lengths (height). Since test data volume and test application time are dominated by the scan tree with the maximum height among the constructed scan trees, the proposed method distributes the scan flip-flops to the scan trees so as to minimize the maximum height of the scan trees. In addition, the proposed method modifies the given test vectors to maximize the reduction in test application time. Experimental results for ISCAS-89 benchmark circuits show that the proposed method could reduce, on the average, test data volume by 77% compared with the conventional multiple scan design. The scan tree construction enlarges the number of scan outputs required. However test data volume could be reduced by 66% even if the number of scan outputs is limited.

Given a test set for stuck-at faults of a combinational circuit or a full-scan sequential circuit, some of the primary input values may be changed to the opposite logic values without losing fault coverage. We can regard such input values as don't care (X). In this paper, we propose a method for identifying the X inputs of test vectors in a given test set. While there are many combinations of X inputs in the test set generally, the proposed method finds one including as many X inputs as possible, by using fault simulation and procedures similar to implication and justification of automatic test pattern generation (ATPG) algorithms. Experimental results for ISCAS benchmark circuits show that approximately 69% of the inputs of uncompacted test sets could be X on the average. Even for highly compacted test sets, the method found that approximately 48% of inputs are X.

This paper presents a method to reduce test data volume and test application time for a full-scan circuit. The proposed method constructs a scan tree in which scan flip-flops are placed and routed in a tree structure. Although one scan input to the scan tree drives several scan chains with varying length, it is guaranteed that every test vector of a test set can be loaded into the scan tree. Since the height of the scan tree decides test data volume of the test set, the method modifies the test set so as to minimize the height. The procedure of test vector modification consists of don't care identification for the test set and a solution to a vertex coloring problem for an incompatibility graph constructed from the test set including don't cares. Experimental results for ISCAS-89 benchmark circuits show that the proposed method could reduce, on average, test data volume and test application time by 70%.

We consider issues related to the reduction of scan test data in designs with multiple scan chains. We propose a metric that can be used to evaluate the effectiveness of procedures for reducing the scan data volume. The metric compares the achieved compression to the compression which is intrinsic to the use of multiple scan chains. We also propose a procedure for modifying a given test set so as to achieve reductions in test data volume assuming a combinational decompressor circuit.

This paper describes a method of test data compression for a given test set using statistical encoding. In order to maximize the effectiveness of statistical encoding, the method first converts some specified input values in the test set to unspecified ones without losing fault coverage, and then reassigns appropriate logic values to the unspecified inputs. Experimental results for ISCAS-89 benchmark circuits show that the proposed method can on the average reduce the test data volume to less than 25% of that required for the original test set.

Given a test set for stuck-at faults, a primary input value may be changed to the opposite logic value without losing fault coverage. One can regard such a value as a don't-care (X). The don't care values can be filled appropriately to achieve test compaction, test data compression, or power reduction during testing. However, these uses are better served if the don't cares can be placed in desired/specific bit positions of the test patterns. In this paper, we present a method for maximally fixing Xs on specific bits of given test vectors. Experimental results on ISCAS benchmark circuits show how the proposed method can increase the number of Xs on specific bits compared with an earlier proposed method.

This paper presents a test vector modification method for reducing power dissipation during test application for a full-scan circuit. The method first identifies a set of don't care (X) inputs of given test vectors, to which either logic value 0 or 1 can be assigned without losing fault coverage. Then, the method reassigns logic values to the X inputs so as to decrease switching activity of the circuit during scan shifting. Experimental results for benchmark circuits show the proposed method could decrease switching activity of a given test set to 48% of the original test set.

We consider issues related to the reduction of scan test data in designs with multiple scan chains. We propose a metric that can be used to evaluate the effectiveness of procedures for reducing the scan data volume. The metric compares the achieved compression to the compression which is intrinsic to the use of multiple scan chains. We also propose a procedure for modifying a given test set so as to achieve reductions in test data volume assuming a combinational decompressor circuit.

Given a test set for stuck at faults, some of primary input values may be changed to opposite logic values without losing fault coverage. We can regard such input values as don't care (X). In this paper, we propose a method for identifying X inputs of test vectors in a given test set. While there are many combinations of X inputs in the test set generally, the proposed method finds one including X inputs as many as possible, by using fault simulation and procedures similar to implication and justification of ATPG algorithms. Experimental results for ISCAS benchmark circuits show that approximately 66% of inputs of un-compacted test sets could be X in average. Even for compacted test sets, the method found that approximately 47% of inputs are X. Finally, we discuss how logic values are reassigned to the identified X inputs where several applications exist to make test vectors more desirable.