James Chien-Mo Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
James Chien-Mo Li has a total of 31 co-authors. Here are the top co-authors by number of shared publications:
- Cheng-Yun Hsieh
Graduate Institute of Electronics Engineering, National Taiwan University, Taipei, Taiwan
3 shared publications Chia-Cheng Pai
Laboratory of Dependable Systems (LaDS), Graduate Institute of Electronics Engineering, National Taiwan University, Taipei, Taiwan
2 shared publications- Tsai-Chieh Chen
Laboratory of Dependable Systems (LaDS), Graduate Institute of Electronics Engineering, National Taiwan University, Taipei, Taiwan
2 shared publications - Tsung-Te Liu
Graduate Institute of Electronics Engineering, National Taiwan University, Taipei, Taiwan
2 shared publications - Akhilesh Kumar
Synopsys, Sunnyvale, CA, USA
1 shared publication Chen-Hung Wu
Graduate Institute of Electronics Engineering, National Taiwan University, Taipei, Taiwan
1 shared publication
Latest Publications
- Open Access
- 20 January 2025
- Published by Association for Computing Machinery (ACM)
- p. 628-634
Efficient ML-Based Transient Thermal Prediction for 3D-ICs
Thermal issues of 3D-ICs have become increasingly severe in recent years. Thus, thermal simulation is needed to ensure thermal safety during the design stage. However, performing thermal simulation iteratively requires a significant amount of time. As a result, a fast and accurate method for thermal prediction is a promising alternative to improve the turnaround time. In this paper, we propose a fast thermal prediction method using machine learning models. In the training phase, we employ two models: one for the initial three time steps and another for the subsequent time steps. To enhance prediction accuracy, we introduce two types of features: spaced-windowed features and time-decayed features. These features help us to capture spatial and temporal information effectively. In our experiment, the mean absolute error for the predicted temperature is 1.12°C, and the maximum error is 7.27 °C. In the prediction phase, we achieve a 116X speed-up compared to a commercial tool. With our proposed method, users can predict transient thermal profiles quickly and accurately to ensure thermal safety.
- 20 January 2025
- Published by Association for Computing Machinery (ACM)
- p. 628-634
Open AccessEfficient ML-Based Transient Thermal Prediction for 3D-ICs
Thermal issues of 3D-ICs have become increasingly severe in recent years. Thus, thermal simulation is needed to ensure thermal safety during the design stage. However, performing thermal simulation iteratively requires a significant amount of time. As a result, a fast and accurate method for thermal prediction is a promising alternative to improve the turnaround time. In this paper, we propose a fast thermal prediction method using machine learning models. In the training phase, we employ two models: one for the initial three time steps and another for the subsequent time steps. To enhance prediction accuracy, we introduce two types of features: spaced-windowed features and time-decayed features. These features help us to capture spatial and temporal information effectively. In our experiment, the mean absolute error for the predicted temperature is 1.12°C, and the maximum error is 7.27 °C. In the prediction phase, we achieve a 116X speed-up compared to a commercial tool. With our proposed method, users can predict transient thermal profiles quickly and accurately to ensure thermal safety.
- Open Access
- 23 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Low-Complexity Algorithmic Test Generation for Neuromorphic Chips
Neuromorphic chips are promising hardware implementations for artificial intelligence (AI) applications owing to their low power consumption. However, neuromorphic chips are difficult to test since they have many potential configurations but lack design for testability (DfT). We propose an algorithmic test generation method for neuromorphic chips without DfT, including fault activation and fault propagation. Fault activation differentiates a neuron's good output and faulty output. Fault propagation sensitizes fault effects to differentiate outputs of faulty chips and good chips. On an L-layer Spiking Neural Network (SNN) model, we achieve 100% fault coverage using O(L) test configurations and test patterns under negligible or no weight variation. Our results show that test effectiveness is maintained even with 4-bit weight quantization. We incur no test escape and overkill even under 10% weight variation. Our total test length is over 73K times shorter than previous works.
- 23 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open AccessLow-Complexity Algorithmic Test Generation for Neuromorphic Chips
Neuromorphic chips are promising hardware implementations for artificial intelligence (AI) applications owing to their low power consumption. However, neuromorphic chips are difficult to test since they have many potential configurations but lack design for testability (DfT). We propose an algorithmic test generation method for neuromorphic chips without DfT, including fault activation and fault propagation. Fault activation differentiates a neuron's good output and faulty output. Fault propagation sensitizes fault effects to differentiate outputs of faulty chips and good chips. On an L-layer Spiking Neural Network (SNN) model, we achieve 100% fault coverage using O(L) test configurations and test patterns under negligible or no weight variation. Our results show that test effectiveness is maintained even with 4-bit weight quantization. We incur no test escape and overkill even under 10% weight variation. Our total test length is over 73K times shorter than previous works.
- 1 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
- Vol. 43 (3) , 826-839
Small Sampling Overhead Error Mitigation for Quantum Circuits
Probabilistic error cancellation (PEC) is a promising error mitigation technique that reduces the error rate without auxiliary qubits. However, PEC has two problems that need to be resolved: (1) There is no good PEC technique for parameterized gates. (2) Sampling overhead grows exponentially with the number of PEC mitigated gates. We first propose a parameterized gate PEC (PGPEC) that mitigates the error without fully characterizing the gates, as the original PEC requires. The result shows that the number of gates requiring characterization for a thousand random circuits can be reduced by 97% or more. We next propose two novel approaches to solving the second problem. We propose a macro gate PEC (MGPEC) technique that aggregates multiple gates as a single macro gate to reduce the exponent of the sampling overhead. MGPEC reduces the sampling overhead by 49% on the QFT7 under the IBMQ noise model, which simulates real operation conditions of quantum circuits. We propose a design diversity PEC (DDPEC) technique to reduce the exponential basis of the sampling overhead. The results show that our DDPEC with design diversity check reduces overall sampling overhead by 13% on the QFT7 circuit under the IBM Q noise model. Combining the DDPEC with the MGPEC, we can reduce overall sampling overhead by 73%.
- 1 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
- Vol. 43 (3) , 826-839
Small Sampling Overhead Error Mitigation for Quantum Circuits
Probabilistic error cancellation (PEC) is a promising error mitigation technique that reduces the error rate without auxiliary qubits. However, PEC has two problems that need to be resolved: (1) There is no good PEC technique for parameterized gates. (2) Sampling overhead grows exponentially with the number of PEC mitigated gates. We first propose a parameterized gate PEC (PGPEC) that mitigates the error without fully characterizing the gates, as the original PEC requires. The result shows that the number of gates requiring characterization for a thousand random circuits can be reduced by 97% or more. We next propose two novel approaches to solving the second problem. We propose a macro gate PEC (MGPEC) technique that aggregates multiple gates as a single macro gate to reduce the exponent of the sampling overhead. MGPEC reduces the sampling overhead by 49% on the QFT7 under the IBMQ noise model, which simulates real operation conditions of quantum circuits. We propose a design diversity PEC (DDPEC) technique to reduce the exponential basis of the sampling overhead. The results show that our DDPEC with design diversity check reduces overall sampling overhead by 13% on the QFT7 circuit under the IBM Q noise model. Combining the DDPEC with the MGPEC, we can reduce overall sampling overhead by 73%.
- 1 August 2021
- Published by Springer Nature in Journal of Electronic Testing
- Vol. 37 (4) , 453-471
Clock-Less DFT and BIST for Dual-Rail Asynchronous Circuits
- 1 August 2021
- Published by Springer Nature in Journal of Electronic Testing
- Vol. 37 (4) , 453-471
Clock-Less DFT and BIST for Dual-Rail Asynchronous Circuits
- 1 November 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-10
qATG: Automatic Test Generation for Quantum Circuits
Researchers now use randomized benchmarking or quantum volume to test quantum circuits (QC) in the laboratory. However, these tests are long and their fault coverage is unclear. In this paper, we propose behavior fault models based on the function of quantum gates. These fault models are scalable because the number of faults is polynomial, not exponential, to the size of QC. We propose a novel test generation that uses gradient descent to generate test configuration with short length. We revise the chi-square statistical method to decide the number of test repetitions under the specified test escape and overkill. Experimental results on IBM Q systems show that our generated test configurations are effective, and our test lengths are 1,000X shorter than traditional test methods.
- 1 November 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-10
qATG: Automatic Test Generation for Quantum Circuits
Researchers now use randomized benchmarking or quantum volume to test quantum circuits (QC) in the laboratory. However, these tests are long and their fault coverage is unclear. In this paper, we propose behavior fault models based on the function of quantum gates. These fault models are scalable because the number of faults is polynomial, not exponential, to the size of QC. We propose a novel test generation that uses gradient descent to generate test configuration with short length. We revise the chi-square statistical method to decide the number of test repetitions under the specified test escape and overkill. Experimental results on IBM Q systems show that our generated test configurations are effective, and our test lengths are 1,000X shorter than traditional test methods.
- 1 September 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 53-58
Diagnosis technique for Clustered Multiple Transition Delay Faults
Power supply noise induced IR drop can cause transition delay faults (TDF) clustered in a small region. However, traditional diagnosis technique cannot handle clustered multiple TDF very well. This paper proposes a diagnosis tool for clustered multiple TDF. Star tracing for TDF is applied to find possible suspects. To tolerate fault masking and fault reinforcement effects, we propose an approximate covering heuristic to find a group of suspects. During approximate covering, we extract important suspects which are likely to be true suspects. We assume many suspects physically cluster around true suspects so our technique prunes suspects based on this assumption. We use correlation coefficient to determine the optimal number of clusters (Optimal NC) so we can apply the K-means algorithm to group suspects. Finally, we prune the least possible cluster but keep important suspects. Simulation on benchmark circuits shows that average accuracy of our tool (0.80) is much better than that of a commercial tool (0.47). Average resolution of our tool (0.35) is also better than that of the commercial tool (0.23).
- 1 September 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 53-58
Diagnosis technique for Clustered Multiple Transition Delay Faults
Power supply noise induced IR drop can cause transition delay faults (TDF) clustered in a small region. However, traditional diagnosis technique cannot handle clustered multiple TDF very well. This paper proposes a diagnosis tool for clustered multiple TDF. Star tracing for TDF is applied to find possible suspects. To tolerate fault masking and fault reinforcement effects, we propose an approximate covering heuristic to find a group of suspects. During approximate covering, we extract important suspects which are likely to be true suspects. We assume many suspects physically cluster around true suspects so our technique prunes suspects based on this assumption. We use correlation coefficient to determine the optimal number of clusters (Optimal NC) so we can apply the K-means algorithm to group suspects. Finally, we prune the least possible cluster but keep important suspects. Simulation on benchmark circuits shows that average accuracy of our tool (0.80) is much better than that of a commercial tool (0.47). Average resolution of our tool (0.35) is also better than that of the commercial tool (0.23).
- 2 February 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
- Vol. 38 (1) , 136-148
DR-Scan: Dual-Rail Asynchronous Scan DfT and ATPG
Due to many state-holding elements in asynchronous circuits, many faults need two-pattern tests. This paper presents a test methodology (DR-scan) for dual-rail asynchronous circuits. Our design for testability (DfT) is a full-scan, clock-less technique that supports both one-pattern and two-pattern tests for single stuck-at faults. DR-scan is able to test memory elements in dual-rail logic without breaking local feedback loops. To reduce test time, we choose a minimum set of selected test configurations. If there are more than one selected test configurations, we need to split scan latches into multiple scan chains. To apply two-pattern tests, we partition the circuit using vertex coloring. With our test methodology, we can apply traditional full-scan automatic test pattern generation (ATPG) to generate two-pattern tests with high test coverage. Experimental results show our methodology can achieve test coverage up to nearly 94% for various asynchronous circuits.
- 2 February 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
- Vol. 38 (1) , 136-148
DR-Scan: Dual-Rail Asynchronous Scan DfT and ATPG
Due to many state-holding elements in asynchronous circuits, many faults need two-pattern tests. This paper presents a test methodology (DR-scan) for dual-rail asynchronous circuits. Our design for testability (DfT) is a full-scan, clock-less technique that supports both one-pattern and two-pattern tests for single stuck-at faults. DR-scan is able to test memory elements in dual-rail logic without breaking local feedback loops. To reduce test time, we choose a minimum set of selected test configurations. If there are more than one selected test configurations, we need to split scan latches into multiple scan chains. To apply two-pattern tests, we partition the circuit using vertex coloring. With our test methodology, we can apply traditional full-scan automatic test pattern generation (ATPG) to generate two-pattern tests with high test coverage. Experimental results show our methodology can achieve test coverage up to nearly 94% for various asynchronous circuits.
- Open Access
- 29 May 2013
- Published by Association for Computing Machinery (ACM)
- Vol. 6, 1-8
GPU-based n-detect transition fault ATPG
This is a massively parallel ATPG that explores device-level, block-level and word-level parallelism in GPU. Eight-detect transition fault ATPG experiments on large benchmark circuits show that our technique achieved 5.6 and 1.6 times speedup compared with a single-core and 8-core CPU commercial tool, respectively. Test patterns selected from our test set are about the same length and quality as those selected from commercial N-detect ATPG. To the best of our knowledge, this is the first proposed GPU-based ATPG algorithm.
- 29 May 2013
- Published by Association for Computing Machinery (ACM)
- Vol. 6, 1-8
Open AccessGPU-based n-detect transition fault ATPG
This is a massively parallel ATPG that explores device-level, block-level and word-level parallelism in GPU. Eight-detect transition fault ATPG experiments on large benchmark circuits show that our technique achieved 5.6 and 1.6 times speedup compared with a single-core and 8-core CPU commercial tool, respectively. Test patterns selected from our test set are about the same length and quality as those selected from commercial N-detect ATPG. To the best of our knowledge, this is the first proposed GPU-based ATPG algorithm.