Neuromorphic architectures are energy efficient architectures for executing spiking neural networks. Current open-source neuromorphic hardware projects are either experimentation platforms (RANC, ODIN) or neural network accelerators (Open-Spike, SNE), there are no direct processing platforms that support AI and ML applications. Seneca-Lite is an open-source RISC-V based multicore neuromorphic platform. The goal of Seneca-Lite is to enable new possibilities for AI and ML applications and foster more collaboration in this field. The platform is intended for research and academic purposes and furthering the field of neuromorphic computing.

The Seneca-Lite platform utilizes the Ibex core, a highly parameterizable open-source 32 bit RISC-V processor. Each core in the Seneca-Lite platform contains a Network On Chip (NoC) router, local memories, network message FIFOs and interconnects. The multi-core platform is designed to facilitate messaging between cores via the NoC. The NoC used in Seneca-Lite is the same as the NoC used in RANC (Reconfigurable Architecture for Neuromorphic Computing). The number of cores in the system is parameterized and can be controlled by the user depending on their need. Since the platform is open-source, many of the internal parameters (Ibex parameters, FIFO parameters etc) can be tweaked by the user as per their target application.

The completed neuromorphic platform is benchmarked for various state-of-the-art applications and compared to other neuromorphic platforms.

One-third of patients suffering from chronic epilepsy, which is caused by abnormal brain activity, is drug-resistant. Animal models are widely used to study the mechanisms leading to epilepsy so better drug treatments can be developed for this disease. In such studies, epileptiform activity, assessed by LFP recordings, can be used as a marker for the development and chronification of disease. However, the analysis of LFP recordings is typically done manually, which is time-consuming, subject to observer bias, error-prone, and lacks consistency and efficiency. Therefore, we present a work which developed a new, automated detection and classification method for epileptiform activity, which was tested in the intrahippocampal kainic acid (IHKA) mouse model, a model of human temporal lobe epilepsy. Our method relies on a spike detector using an improved version of the nonlinear energy operator (NEO) in combination with automatic NEO thresholding (ANT). The detected spikes form the basis of epileptiform event detection and classification. The proposed method is implemented in Python as an automated and time-efficient algorithm that can be used in preclinical studies. Epileptiform event detection accuracy was 93.1% and classification accuracy 95.8%. Moreover, the time for analysis of LFP recordings was reduced by 98.8% compared to manual analysis. Additionally, to demonstrate the potential of the algorithm for application in Brain-Machine Interfaces (BMI), we performed a real-time implementation using both an application-specific integrated circuit (ASIC) and a field programmable gate array (FPGA). The FPGA demonstrated the feasibility of real-time implementation, and the ASIC resulted in an area and power efficient chip using the Taiwan semiconductor manufacturing company (TSMC) 45nm library that constitutes a post-layout area of 9114 µm2 and a power usage of 6.11 µW.

As technology nodes continue to shrink, more challenges arise in the field of Design for Testability (DfT). Sequential Integrated Circuits (IC) with asynchronous (re)set flip-flops are notorious for producing unwanted reset behaviour during scan-test. Typically the scan flip-flops are restricted to a non-reset state while the shift operation is performed. This can be achieved by inserting an independent test signal in parallel with each local reset port. This ensures that the scan flip-flops are loaded with the correct input data during the shift cycle. However, before the capture cycle is initiated, this test signal must be released as it prohibits the reset logic from being tested with stuck-at-faults. When there are multiple cascading resets present in the design this release can cause glitches to occur. Since the reset ports operate asynchronously, these glitches can also trigger a scan flip-flop to reset, thereby changing the input data. As a result, some chips may be tested with corrupted input data, leading to differences in scan patterns. Consequently, these ICs fail during manufacturing tests and are classified as faulty, leading to yield loss.


By adding DfT to the (re)set port of each flip-flop these glitches can be prevented at the cost of additional hardware. This thesis establishes the conditions that lead to the occurrence of asynchronous (re)set glitches during scan-test. A design rule-based algorithm is proposed that can accurately identify glitchy structures for circuits without reconvergence. As an extension, a simulation-based algorithm is presented that can further classify which (re)set flip-flops can cause a glitch. These algorithms have been tested on two case studies where glitches have been observed. After deploying these algorithms 33% and 71% of the total number of (re)set flip-flops were identified as glitch-free. By only adding additional DfT to these flip-flops the overhead is significantly reduced. This work addresses a significant challenge in minimizing the cost of robust asynchronous scan test.

Deep Neural Networks (DNNs) have revolutionized numerous computational fields, from image and speech recognition to autonomous driving and natural language processing. Yet, the substantial computational and energy requirements of DNNs, particularly Convolutional Neural Networks (CNNs), pose significant obstacles to their deployment on resource-constrained edge devices. This thesis presents SPARO, a novel Scalable Sparsity-Aware Event-Driven Architecture designed to overcome these challenges by effectively exploiting sparsity in both neural network weights and activations.
SPARO’s architecture is founded upon a unique event-driven dataflow that harnesses the inherent sparsity of CNNs, thereby reducing computational burden and energy consumption. This dataflow is strategically divided into two distinct phases: the Update Phase and the Fire Phase. During the Update Phase, all computations essential for incoming events are executed, while the Fire Phase is dedicated to applying non-linear activation functions and pooling operations to the output feature maps (OFM). This meticulously designed phased approach streamlines data handling, eliminates redundant computations, and significantly boosts overall processing efficiency.
A cornerstone of SPARO’s innovation is its dynamic weight reuse mechanism, which intelligently maximizes the reuse of weights across multiple events. This significantly reduces the number of weight fetches needed, thereby improving arithmetic intensity. Furthermore, SPARO leverages advanced sparse data representation techniques to minimize memory usage and further enhance computational efficiency.
The efficacy of SPARO is demonstrated through comprehensive evaluations using both synthetic benchmarks and real-world CNN applications, such as gesture recognition and object detection. In the same form-factor, SPARO achieves an impressive 8.5x speedup compared to the baseline Seneca system, delivering real-time performance while consuming only 14% of the energy for the TinyYolo vision task.

Memory advances have not kept up with computing demands. Emerging device technology Resistive RAM (RRAM) addresses this by enabling computation-in-memory. However, RRAM suffers from read disturb, limiting viability. While earlier work has had some success in reducing read disturb by switching the read current direction (a bipolar read scheme), the RRAM device eventually degraded. In this work, a reinforcing bipolar read scheme is introduced, which aims to prevent read disturb by reinforcing both HRS and LRS, away from the undefined state. From device-level simulation, this reinforcing behaviour is predicted for V = 0.5, 0.4 and 0.3 V with a fixed switching ratio between time under positive and negative read voltage polarity rsw i tch = 2.5, 2.5 and 2.75, respectively. At V = 0.2 V and V = 0.1 V, it is predicted that no reinforcement scheme exists. The bipolar read scheme found for V = 0.5 V was evaluated in circuit simulation with a sense amplifier. however the HRS and LRS reinforcing behaviour found at device-level could not be replicated. From unipolar read results at circuit level, it was determined no switching ratio exists that reinforces both the HRS and LRS state boundaries as chosen here at RHRS,min = 12.8kΩ and RLRS,max = 4.4kΩ for V = 0.5 V. However, a 5.1x reduction in resistance drift compared to conventional unipolar read was still obtained.

Motivated by the desire to bring intelligent processing at the Edge, enabling online learning on resource- and latency-constrained embedded devices has become increasingly appealing, as it has the potential to tackle a wide range of challenges: on the one hand, it can deal with on-the-fly adaptation to fast sensor-generated streams of data under changing environments and on the other hand, it can address a variety of challenges associated with offline training in the cloud, such as incurred energy consumption of sensor data transfers and extra memory storage for the training samples, but also data privacy and security concerns. Concurrently, maintaining low-latency and power-efficient inference is paramount for edge AI computing systems, and thus learning/adapting online with minimal incurred overhead is crucial.

In this work, we propose EON-1, an Edge ONline Learning SCNN (Spiking Convolutional Neural Network) processor with 1-bit synaptic weights, 1-spike per neuron and 1-neuron updated per input, which we have benchmarked for both ASIC and FPGA platforms. Our key contribution is proposing a binary and stochastic SDTP rule which, benchmarked in an ASIC node, achieves less than 1% energy overhead for inference. To our knowledge, our solution incurs the least energy overhead for inference, compared to state-of-the-art solutions, showing a better efficiency by at least a factor of 10x. We also report 94% and 77.65% accuracy on the MNIST and Fashion-MNIST classification tasks, and we achieve 0.09pJ/SOP and 1.5pJ/SOP energy efficiency during inference and learning, respectively. We extend our solution to demonstrate a practical use-case of performing inference in real-time UHD videos while coping with streaming data and we showcase 60 FPS UHD video processing.

Modern computer application require large amounts of data processing. Traditional computing models involve constant data transfer between memory and processor. This data transfer is a major contributor to high energy consumption. As these applications scale, the energy demand increases. This poses challenges in terms of sustainability and operational costs. Computation In Memory (CIM) integrates processing within the memory. This reduces the need for data transfer between memory and processor. Potential for drastically lowering energy consumption.

CIM macros are often implemented using modified SRAM cells, though recent literature explores memristor-based CIM designs due to the memristor’s low-energy, non-volatile characteristics. However, no comprehensive comparisons between SRAM-based and memristor-based CIM designs exist. While memristor-based designs are hypothesized to be more energy-efficient, this has not yet been proven.

This thesis compares SRAM-based and memristor-based CIM designs to determine which is better suited for CIM applications. This has been achieved by exploring the state of the art of memristive devices, memristor based CIM macros and SRAM based CIM macros. A selection of designs were chosen to compare, including the 1T1R and 8T SRAM design, which are the most popular memristor based and SRAM based CIM designs. The schematics of all the designs were recreated and simulated using as much of the same parameters as possible in all of the designs. A simulation of performing the logic AND and the MAC operation was made. Additionally a layout of the designs was made to extract the area. The designs were compared based on area, energy consumption and delay.

From the results could be concluded that the best device for CIM depends on the application. The memristor design had the smallest area and consumed the least amount of energy for reading, logic and MAC operations. The memristor design also consumed the most amount of energy during writing and the delay for all operations is longer than with the SRAM based designs. If area, energy consumption and delay are equally important for an application, then memristor based CIM would be the better choice only if there are much more logic/read operations than write operations. It could be the better choice for MAC operations if a more energy efficient ADC was used than the one used in this thesis.

Machine learning on edge devices performs crucial identification or prediction tasks while limiting the amount of data that needs to be transmitted to more centralized computing nodes. However, strict area and energy requirements necessitate specialized hardware developed for the requirements of the device and model. This thesis is concerned with developing an area and energy arithmetic unit as part of the implementation of a stacked machine learning model in embedded automotive devices. The model in question was previously designed to perform lifetime prediction with the goal of improving the reliability of semiconductor devices used in various automotive applications.

This thesis aims to achieve area and energy efficiency by exploiting the commonalities in the arithmetic operations of several of the internal learners of the stacked machine learning model. The use of a weighted figure of merit, taking into account area, energy and delay, allow for simple comparisons of designs at any operation frequency and easy insight into the changes in the merit of designs if device requirements were to change. A sweep of the percentage of multiplications in the workload also gave insight into how design choices may change due to future redesigns of the stacked machine learning model.

It was found that the MAC, multiply, divide and accumulate operations of the internal learners can best be supported by one arithmetic unit containing a "Reduced Area" parallel multiplier (still taking up most of the area), a small, dedicated accumulator and invariant integer division using the multiplier. It was also found that the ability to reconfigure the multiplier for different levels of bit-precision does not yield performance improvement for the expected precision distribution.

The security of electronic devices holds the greatest importance in the modern digital era, with one of the emerging challenges being the widespread occurrence of hardware attacks. The aforementioned attacks present a substantial risk to hardware devices, and it is of utmost importance to comprehend the potential detrimental effects they may cause. Side-channel attacks are a class of hardware attacks that exploit information unintentionally leaked by a device during its operation. These leaks manifest in various forms, including power consumption, time variations, and thermal dissipation. The fundamental danger posed by side-channel attacks is their ability to infer sensitive information from these unintended emissions. To address the heightened risks associated with side-channel attacks, this thesis focuses on three main research topics.@en

Conventional computing systems involve physically separated storing and processing units. To perform the processing, data is shuttled from the storing unit to the processing unit followed by the actual processing, and the processed data is shuttled back into the storing unit. Unfortunately, this data shuffling contributes significantly to the overall latency and energy consumption of the system. Computation-in-memory (CIM) offers a promising alternative that can process the data within the storing unit, thereby, alleviating the need for data shuffling. This can potentially lead to high energy efficiency and high throughput computation. In addition, emerging non-volatile memristors provide an excellent storing element that can inherently perform the computation in CIM while also retaining the data value.
Memristor-based CIM has been vastly explored to perform certain data-intensive computational tasks for numerous applications related to artificial intelligence (AI), Big Data, and data encryption while realizing high-performing, low-power micro-architectural solutions. This thesis first identifies the existing key challenges related to emerging memory technology and CIM memory array, and the circuit design of periphery logic to achieve low-power and high-performing CIM units. Thereafter, it presents several micro-architectural solutions to build CIM accelerators that can perform logic and arithmetic operations in an efficient manner.

Identifying the challenges of memristor-based CIM: The thesis briefly summarizes the key aspects of a memristor-based CIM architecture that can perform certain logic and arithmetic operations. First, the key components of a typical CIM architecture are presented, highlighting the design considerations to build these components. An account of computational accuracy and efficiency share of each of these components is presented supported by a comprehensive literature survey while highlighting the underlying concepts of each component. This is followed by identifying the key challenges in achieving stringent efficiency metrics required by the targeted applications while dealing with non-idealities. Finally, the limitations of the state-of-the-art solutions that target these challenges are highlighted, thus, providing the motivation to develop outperforming solutions that are presented in the thesis.

Developing micro-architectural solutions for efficient CIM-based logic accelerators: This part of the thesis presents several logic accelerators that can perform (N)OR, (N)AND, and XOR operations in a fast and energy-efficient manner. \textbf{FIVE} different solutions target different aspects of performing CIM-based logic, whereby, scaling the number of operands per cycle and the memory size, and expanding to different types of operations that can be supported while addressing the non-idealities. State-of-the-art solutions are comprehensively outperformed by our proposed solutions in terms of energy efficiency, performance, and the maximum number of operands that can be accurately performed in a single cycle.

Developing micro-architectural solutions for efficient CIM-based arithmetic accelerators: This part of the thesis presents several compact arithmetic accelerators that can perform multiply-and-accumulate (MAC) operations with low energy consumption. \textbf{THREE} different analog-to-digital converter (ADC) topologies are presented with two of them demonstrated with a chip prototype. Optimizing the memory structure is also explored to ensure accurate generation of analog output in the CIM memory array and to optimize the A/D conversion. A comparison of the proposed solutions with the state-of-the-art is presented, highlighting the promise of the developed micro-architectures.@en

Artificial intelligence (AI) is rapidly becoming an integral part of many real-world products and services. This is mainly facilitated by the extensive computing resources provided by the cloud infrastructure. However, cloud-based AI processing suffers from drawbacks like high latency, huge network costs, data privacy/security concerns, and service disruptions due to internet outage. Edge computing for AI (edge-AI) addresses these problems by combining data sources with on-board AI processing hardware. Such hardware must be energy efficient to achieve prolonged operation, given the limited energy resources on edge devices. Moreover, it should be compact in size to facilitate seamless system integration and enhanced portability. Conventional hardware cannot meet these requirements due to data transfer bottleneck in von Neumann architecture and limitations of conventional memory technologies.
Computation-in-memory (CIM) overcomes these challenges by in-situ data processing using emerging memory technologies called memristors. Thus, CIM can facilitate energy efficient and compact edge-AI hardware design. Healthcare domain stands out as a prime target for CIM-based edge-AI hardware, due to two main reasons. Firstly, it holds significant real-world importance due to its direct impact on human well-being. Secondly, the increasing adoption of AI in healthcare can significantly benefit from efficient hardware for data processing. CIM-based edge hardware can greatly enhance the effectiveness of AI-based healthcare through rapid, reliable, and secure processing of medical data at its source. Hence, design of CIM-based edge-AI hardware for healthcare applications presents a promising research direction.

The process of designing CIM-based edge-AI hardware for healthcare can be expressed as a stack of six abstraction layers: application, algorithm, optimization, mapping, micro-architecture and circuits, and device. These abstraction layers can be further grouped into two distinct design phases. The first phase is application-dependent, covering the first three abstraction layers (application, algorithm and optimization). It involves creating a customized neural network model for the given healthcare application. The challenge in this phase is to achieve strong algorithmic performance, while incorporating features to exploit the full potential of CIM hardware. Conversely, the second phase is application-independent and comprises of the remaining abstraction layers (mapping, micro-architecture and circuits, and device). It solely focuses on translating the model computations into CIM hardware operations. However, the non-ideal characteristics of memristor devices introduce computational errors in hardware operations. This undermines the advantages of CIM as energy-efficient computations are of no use if they are incorrect. Hence, mitigating memristor non-idealities becomes the primary challenge in this phase. Moreover, it is important to integrate the customized model and non-ideality mitigation strategies into a comprehensive hardware solution and realize it through prototyping. This gives rise to the following three research topics: 1) healthcare AI models for CIM-based edge hardware, 2) dealing with memristor non-idealities, and 3) CIM edge-AI prototyping for healthcare.

We adopt a cross-layer approach in this thesis to address these research topics, covering all six layers of the CIM abstraction stack. We begin by creating neural network models for two healthcare applications: cardiac arrhythmia classification and diabetic retinopathy screening. Our contributions in this application-dependent design phase span across the first three abstraction layers (application, algorithm and optimization). At the application layer, we introduce new features in the model tailored to the specific healthcare application. This enhances its real-world impact by addressing the unique medical needs more effectively. Moving to the algorithm layer, we customize the computational flow within the model to exploit the characteristics of the healthcare data. This improves design performance in key aspects like accuracy and energy efficiency. Moreover, we strategically refine the model computations to further maximize post-deployment benefits on CIM hardware. At the optimization layer, we employ techniques like resampling, quantization and pruning to optimize hardware resource requirements, without compromising the model's algorithmic performance.

After creating the neural network models, we proceed to the application-independent design phase. Focusing on RRAM-based memristor devices, we first identify three key non-idealities that significantly impact inference accuracy on CIM hardware. We then devise mitigation strategies against these non-idealities, encompassing the remaining abstraction layers (mapping, micro-architecture and circuits, and device). At mapping layer, we propose a hardware-aware training methodology to combat the conductance variation non-ideality. Moving to the micro-architecture level, we present two mitigation strategies. The first addresses non-zero Gmin error non-ideality through a novel approach to CIM micro-architecture design. The second introduces an adaptive micro-architecture that adjusts its sensing conditions to counteract the effects of read-disturb non-ideality. At the device level, these strategies indirectly contribute by circumventing the necessity for extensive device engineering, ensuring accurate inference even in the presence of non-idealities. Building upon this foundation of model development and non-ideality mitigation, we integrate the optimal ECG classification model with the proposed mitigation strategies to create a CIM edge-AI prototype. Thus, our contributions pave the way towards a future with enhanced effectiveness and efficiency of AI-powered healthcare.@en

Cardiovascular diseases (CVDs) are the top cause of death worldwide, and their diagnosis can be quickly and painlessly achieved through Electrocardiogram (ECG). The diagnosis of electrocardiogram has gradually evolved from manual diagnosis by doctors to one that can be realized using Artificial Intelligence (AI). Early AI still required manual extraction of features for ECG classification, but later Deep Neural Network (DNN) could automatically extract features during the learning process. With this technology, people can monitor heart movements in real-time through wearable devices. If there are any abnormalities, they can seek medical treatment in time to prevent death from sudden severe heart disease. However, for the wearable device, the energy consumption per classification by the traditional AI is so high that a limited battery cannot work for a long time.

To address this issue, this thesis adopts Spiking Neural Network (SNN) to do classification and implement the inference on the hardware. Compared with traditional Artificial Neural Network (ANN), SNN is highly energy efficient. According to the needs of SNN and its event-driven characteristics, the multi-threshold-based encoding scheme is proposed, which encodes the heartbeat into 54 spikes on average with less information loss. The SNN model is trained by ANN-SNN conversion with an accuracy of 97.42\%. After RTL coding, synthesis, and back-end implementation, the chip with encoding and inference functions achieves an energy consumption of only 57.88 nJ per heartbeat classification.

In 2022, there were over 26 million electric automobiles on the road, a 60% increase with regard to 2021 and more than 5 times the stock in 2018. As automobiles become more electric and systems get increasingly complex, the safety requirements get more stringent. In 2011, the International Organization for Standardization (ISO) established ISO26262 to provide guidance to the semiconductor industry on the development process of safety essential ICs for automotive applications.

When evaluating the safety of a system parallel to the normal functional operation, traditional Design-for-Test (DFT) techniques such as scan chain, Built-In Self-Test (BIST), Joint Test Action Group (JTAG), and boundary scan are no longer viable options for two fundamental reasons. Firstly, safety assessment in the context of automotive applications necessitates evaluation at an application level, going beyond the capabilities of these techniques, which are primarily designed for structural and functional testing. Secondly, during safety checks, it is imperative that the normal operation of the integrated circuit (IC) remains uninterrupted, as the chip is often deployed in critical, real-time systems. These traditional DFT methods, while effective during the IC manufacture and production before they are released in the market, fall short of addressing the dynamic and application-specific safety concerns that arise during the operational lifecycle of safety-critical systems in sectors like automotive engineering.

To address these challenges, fault injection has emerged as the necessary step for safety assessment. ISO26262 explicitly recognizes fault injection as one of the most popular techniques for evaluating a system's safety and determining its Automotive Safety Integrity Level (ASIL). The safety metrics are required for certification of a product with ASIL. Fault injection allows for the creation of realistic fault scenarios and the assessment of how the system responds to these faults during normal operation, aligning more closely with the dynamic and application-specific nature of safety-critical systems in fields of automotive engineering.

Currently, many EDA companies provide Failure Mode Effects and Diagnostic Analysis (FMEDA) platforms for IC safety evaluations. However, these tools are time-consuming and resource-intensive. Additionally, as IC designs become more intricate, there is a reliance on fault reduction techniques such as statistical sampling, which entails simulating only a mere 5% of the overall fault space. Thus, an FPGA-based fault emulation system emerges as a promising approach to expedite this process.

The novelty of this work was in designing a dedicated platform tailored for the evaluation of safety-critical systems for automotive applications such as Battery Management Systems (BMS). This platform can execute the safety sequences on the system to evaluate the safety mechanisms implemented in the design in the presence of random faults assuming the chip is in use in the car. Moreover, the fault emulation activity provides the evidence necessary for the certification of products with ASIL level. Furthermore, performing this activity during the development stage helps in designing the ICs with the highest level of safety. The proposed FPGA-based fault emulation system efficiently overcomes three key challenges: it decreases execution time dramatically provides a speed-up of 296x compared to the simulation method, optimizes resource utilization, eliminates the tool license cost, and removes the requirement for considerable fault space reduction. This platform can emulate a large fault population of up to one million faults in less than three hours.

Computer Architects often walk the tightrope between performance, power and area while designing modern day processors. This daunting task is made even more challenging by short Time-to-Market requirements set by the clients. In light of these challenges, architectural simulators provide a much needed tool for the architects to gauge the impact of their innovations rather quickly. Arguably, use of such simulators is essential in avoiding a product recall due to the processor failing to deliver on performance/power requirements for the intended application. The thesis is in-line with the objective of identifying and addressing the bottlenecks in the RISC-V simulation ecosystem and contribute in the development of RISC-V infrastructure.
The intended objective of an architectural simulator is to capture the trend of the real hardware (i.e., performance improvement due to micro-architectural changes in real hardware should be eloquently captured by the simulator). A good simulator shall have high throughput (less simulation time) and should be easily re-configurable. The re-configurability of the simulator can be as fine as micro-architectural changes or as large as a new ISA being simulated. These attributes of speed and re-configurability come at the cost of accuracy. A high error in performance statistics of the simulator fails to engender confidence among the prospective users. Hence, validating performance of simulators against hardware is essential.
The thesis introduces the need for a full system architectural simulator for RISC-V processors followed by a brief, yet crisp review of the past attempts at making such simulators. The review is from the perspective of existing methodologies for performance validation of the simulators. The work also proposes a new methodology for validating system simulators. Although, the proposed methodology is generic and can be extended to other ISAs (such as ARM, x86, etc.), the target hardware chosen are RISC-V ISA based systems that span both commercially, IP protected processor as well as open-source processors widely adopted by the RISC-V community.
The error in validated simulator is reduced to 22.9% and 18.9% for selected SPEC2017 benchmarks applications, by calibrating the CPU model. The methodology also streamlines CPU performance validation of the simulators.

Purkinje cell is a type of neuron that can be found in the cerebellum. What characterises Purkinje cell neural activity is the fact that it exhibits two types of spiking behaviour; the so-called simple and complex spikes. These two types of spikes are thought to play a role in motor functionality. In order to better understand the relationship between Purkinje cell neural activity and the motor-cortex, neuroscientists record such neural activity in mice. However, current experimental setups pose a challenge as they involve a wired connection between the animal’s head stage and the recording device, which limits the mouse’s natural behaviour by restricting its movement. This work proposes a lightweight neural-spike detection and classification architecture for acquiring Purkinje cell neural activity. The proposed design discards unneeded information, by detecting and classifying spikes in real-time. This type of compression enables data storage on a removable device in the head stage, freeing mice from wires. Its small formfactor allows unrestricted movement during experiments, while a power-efficient design ensures long-termoperation. The performance of the algorithm has been evaluated using a software implementation, yielding a combined accuracy for detection and classification ranging from 92.74% to 94.54%. The system has been synthesised using the 45 nm Nangate Open Cell library resulting in an ASIC with an area of 0.22mm2 and a power consumption of 0.412mW.

This thesis focuses on identifying and classifying defects in STT-MRAM technology using novel and machine learning approaches. The thesis discusses the basic principles of STT-MRAM and the semiconductor chip manufacturing process and test stages. The research aims to develop novel methods and explore machine-learning approaches to diagnose defects in STT-MRAM devices. The current defect identification methodologies have shown certain cost, speed, and scalability limitations. The thesis presents DAT-based and ML-based Diagnosis methodologies to identify and classify STT-MRAM unique defects to address these challenges. The methods are evaluated and validated on experimental wafers performed at IMEC in Leuven, Belgium.

DAT-based Diagnosis involves automated defect identification in STT-MRAM based on identifying features automatically extracted from specialized measurements targeting the unique defects, Pinhole, Intermediate State, SAF Flip, and Back-Hopping. ML-based Diagnosis uses machine learning techniques to classify defects using MTJ features extracted from low-cost measurements. Data collected from electrical measurements on experimental STT-MRAM devices serve as the basis for evaluating the developed methodologies. The thesis also discusses data analysis, including data visualization, feature correlations, and outlier analysis for future research. Furthermore, a machine learning training process is performed, including hyperparameter optimization and evaluation using F-score and B-accuracy metrics to assess the model's performance and the ability to generalize on unseen data.

DAT-based Diagnosis aims to maximize the defect detection accuracy at the expense of measurement costs. In contrast, ML-based Diagnosis minimizes the measurement cost while maximizing the detection accuracy for robust and balanced classification. However, the DAT-based Diagnosis is not verified using PFA to validate the defect types identified by the developed methodology. Furthermore, the ML-based Diagnosis uses training data labeled by the unverified DAT-based Diagnosis approach to train machine learning models. Despite these limitations, the results have shown valuable insights into defect identification and classification, proving a robust framework for diagnosing STT-MRAM devices. Additionally, a scientific paper is submitted on march-based diagnosis, adapting the DAT-based Diagnosis method to industrial chips that are limited in extracting the identifying features.

Resistive random access memory (RRAM) is an emerging memory technology that has the potential to replace dynamic random access memory (DRAM) or FLASH. The current memory technology suffer from scalability issues. RRAM can be used as potential replacement for Flash and DRAM. RRAM stores information using resistance states instead of charge. RRAM is non-volatile memory, power efficient, scalable and compatible with the current CMOS process.
Before RRAM can be commercialized the quality of RRAM devices needs to be guaranteed. For this we need to diagnose RRAM devices. The diagnosis allows us to improve the manufacturing process as well as built defect models for memory testing. The traditional fault models do not incorporate the non-linear behaviour of the RRAM device. As a result, tests are created for the wrong test space resulting in a lower yield and more test escapes. To increase the yield and create more reliable fault models, defect models that incorporate the physical defect of the RRAM device are required. To ensure the manufacturing quality of the RRAM devices and for memory testing of RRAM the characterization and diagnosis of RRAM is required. The electrical characterization of RRAM is cheap, fast and is used to evaluate the performance of RRAM. For diagnosis electrical characterization has not been used before.
This work uses the electrical characterization of RRAM to automatically identify defective RRAM devices which can be used for diagnosis. For the defect identification, the key parameters are determined. To assess the performance of the defect identification algorithms the RRAM devices are labelled manually. In total five methods have been implemented to automatically label defective RRAM devices. A statistical analysis is performed on the RRAM devices. Using the labelled data the devices are compared to a nominal device to identify defects in RRAM. Furthermore, K-means, KNN and a CNN algorithm is applied to RRAM. The classification algorithms allow the automatic identification of defective RRAM devices which can be used for diagnosis.
The metrics used for the statistical analysis are insufficient to accurately identify defective RRAM devices. The nominal device method classifies 81% of the device correctly using the euclidean algorithm. The best performance is obtained for the supervised learning algorithm. K-NN classifies 94% of the cycles correct and 84% of the devices. If the data is not labelled the unsupervised learning algorithm can be used. Kmeans classifies 79.6% of the devices correctly which is slightly worse than K-NN. The CNN classifies 67% of the devices correctly using 20 epochs. However, the CNN has not yet been optimized and needs to be improved. For unlabelled data, the unsupervised learning algorithm should be used and for labelled data the K-NN. The electrical characterization of RRAM devices using machine learning looks promising and is much cheaper compared to optical characterization and memory tests. If in the future the defective devices can be linked to the underlying defect, this will lead to cheaper diagnosis of RRAM devices and allow the creation of accurate fault models.

3D NAND memory devices are intrinsically very cost sensitive, implying that their size, and hence logic area must be limited in order to acquire a chip which is able to conquer the competitive market price. Market forecasts of upcoming NAND products predict Input/Output (I/O) speed increase well beyond 2 Gb/s that is the current industry standard. I/O bandwidth strongly correlates with the device technology used for logic and the current state-of-the-art planar devices are predicted to reach their maximum capabilities in the near future. Use of FinFET transistor is expected to substantially enhance I/O area-performance of 3D NAND memory logic, alleviating area restriction severity and providing a foundation for future periphery generations to come.

In this thesis, a 3D NAND compatible I/O able to achieve 8 Gb/s throughput has been developed using simulation of thermally stable Imec in-house developed 14 nm FinFET technology equivalent. To validate throughput quality, industry defined eye diagram standards are used. To determine area savings provided by utilizing FinFET devices, FinFET active transmitter area is benchmarked against 45 nm planar device setup achieving the same 8 Gb/s data rate performance. To ensure an unbiased comparison, two signaling topologies are used - single ended signaling (SES) and differential signaling (DS). To extend analysis, sensitivity of the design against various parameters such as data rate, voltage and temperature is explored.

It is concluded, that active area of FinFET driver is several times lower than that of similar planar transmitter (same power and throughput) for both SES and DS. Additionally, suitable use cases of DS and SES have been evaluated depending on environmental conditions investigated during sensitivity analysis. All in all, this research provides a baseline for planar-to-FinFET scaling in I/O system and guidelines in choosing signaling topology appropriately, depending on system constraints.

Resistive Random-Access Memory (RRAM) is an emerging memory technology that has the possibility to compete with mainstream memory technologies such as Dynamic Random-Access Memory (DRAM) and flash memory. The reason why RRAM has not seen mass adoption yet is due to its defect-prone nature. The resistance of RRAM can assume any value within its operating range and its resistance can be divided into five states instead of the regular two logic states. Conventional test techniques are incapable of detecting unique faults due to their inability to distinguish between all five cell states, resulting in a large number of test escapes. Therefore, new test methods, such as Design-For-Testability (DFT), need to be developed to reduce the number of test escapes and ensure customer satisfaction. This work proposes two new DFTs: Parallel-Reference Read (PRR) and Closed-Loop Write (CLW). The PRR DFT is a replacement for the regular read circuit, which enables the detection of all five cell states, while the CLW DFT is an addition to the regular write circuit, which introduces feedback during the write operation. From these two DFTs, the PRR DFT is selected for further development and its design is validated. From the validation, it is concluded that the PRR DFT can detect all five cell states. Moreover, under process variations, the PRR DFT will provide the correct output in 95.90% of the cases. Furthermore, the PRR DFT improves the overall resistive-defect detection capability by 14.79% when compared to a regular read circuit. Finally, the PRR DFT offers 100% identified fault coverage while only requiring 4N write operations, 5N read operations and an area overhead of 14N_c transistors, where N and N_c are the total number of cells and the total number of columns in the RRAM array, respectively.