Search

Large language models (LLMs) have catalyzed an upsurge in automatic code generation, garnering significant attention for register transfer level (RTL) code generation. Despite the potential of RTL code generation with natural language, it remains error-prone and limited to relatively small modules because of the substantial semantic gap between natural language expressions and hardware design intent. In response to the limitations, we propose a methodology that reduces the semantic gaps by utilizing C/C++ for generating hardware designs via High-Level Synthesis (HLS) tools. Basically, we build a set of C-to-HLS optimization strategies catering to various code patterns, such as nested loops and local arrays. Then, we apply these strategies to sequential C/C++ code through in-context learning, which provides the LLMs with exemplary C/C++ to HLS prompts. With this approach, HLS designs can be generated effectively. Since LLMs still face problems in determining the optimized pragma parameters precisely, we have a design space exploration (DSE) tool integrated for pragma parameter tuning. Furthermore, we also employ profiling tools to pinpoint the performance bottlenecks within a program and selectively convert bottleneck components to HLS code for hardware acceleration. By combining the LLM-based profiling, C/C++ to HLS translation, and DSE, we have established HLSPilot, the first LLM-enabled high-level synthesis framework, which can fully automate the high-level application acceleration on hybrid CPU-FPGA architectures. According to our experiments on real-world application benchmarks, HLSPilot achieve comparable performance in general and can even outperform manually crafted counterparts, thereby underscoring the substantial promise of LLM-assisted hardware designs.

Mixed-precision neural network (MPNN) that utilizes just enough data width for the neural network processing is an effective approach to meet the stringent resources constraints including memory and computing of MCUs. Nevertheless, there is still a lack of sub-byte and mixed-precision SIMD operations in MCU-class ISA and the limited computing capability of MCUs remains underutilized, which further aggravates the computing bound encountered in neural network processing. As a result, the benefits of MPNNs cannot be fully unleashed. In this work, we propose to pack multiple low-bitwidth arithmetic operations within a single instruction multiple data (SIMD) instructions in typical MCUs, and then develop an efficient convolution operator by exploring both the data parallelism and computing parallelism in convolution along with the proposed SIMD packing. Finally, we further leverage Neural Architecture Search (NAS) to build a HW/SW co-designed MPNN design framework, namely MCU-MixQ. This framework can optimize both the MPNN quantization and MPNN implementation efficiency, striking an optimized balance between neural network performance and accuracy. According to our experiment results, MCU-MixQ achieves 2.1$\times$ and 1.4$\times$ speedup over CMix-NN and MCUNet respectively under the same resource constraints.

FPGA-based graph processing accelerators, enabling extensive customization, have demonstrated significant energy efficiency over general computing engines like CPUs and GPUs. Nonetheless, customizing accelerators to diverse graph processing algorithms with distinct computational patterns remains challenging and error-prone for high-level application users. To this end, template-based approaches have been developed to automate the graph processing accelerator generation. Although these frameworks significantly enhance the design productivity, the templates often result in closely coupled algorithms, programming models, and architectures, severely limiting the versatility of the targeted graph processing algorithms and their applicability to high-level users. Furthermore, the limitations of the frameworks are usually ambiguous due to the absence of a rigorous grammar definition. To overcome these challenges, we introduce Graphitron, a domain-specific language (DSL), which allows users to generate customized accelerators for a wide range of graph processing algorithms on FPGAs without engaging with the complexities of low-level FPGA designs. Graphitron, by defining vertices and edges as primitive data types, naturally facilitates the description of graph algorithms using edge-centric or vertex-centric programming models. The Graphitron back-end employs a suite of hardware optimization techniques including pipelining, data shuffling, and memory access optimization that are independent with the specific algorithms, supporting the creation of versatile graph processing accelerators. Our experiments indicate that accelerators crafted using Graphitron achieve comparable performance to that generated with template-based design framework. Moreover, it exhibits exceptional flexibility in algorithm expression and significantly enhance accelerator design productivity.

Open-source EDA tools are rapidly advancing, fostering collaboration, innovation, and knowledge sharing within the EDA community. However, the growing complexity of these tools, characterized by numerous design parameters and heuristics, poses a significant barrier to their widespread adoption. This complexity is particularly pronounced in integrated circuit (IC) backend designs, which place substantial demands on engineers' expertise in EDA tools. To tackle this challenge, we introduce IICPilot, an intelligent IC backend design system based on LLM technology. IICPilot automates various backend design procedures, including script generation, EDA tool invocation, design space exploration of EDA parameters, container-based computing resource allocation, and exception management. By automating these tasks, IICPilot significantly lowers the barrier to entry for open-source EDA tools. Specifically, IICPilot utilizes LangChain's multi-agent framework to efficiently handle distinct design tasks, enabling flexible enhancements independently. Moreover, IICPilot separates the backend design workflow from specific open-source EDA tools through a unified EDA calling interface. This approach allows seamless integration with different open-source EDA tools like OpenROAD and iEDA, streamlining the backend design and optimization across the EDA tools., Comment: under review

Deep learning and signal processing are closely correlated in many IoT scenarios such as anomaly detection to empower intelligence of things. Many IoT processors utilize digital signal processors (DSPs) for signal processing and build deep learning frameworks on this basis. While deep learning is usually much more computing-intensive than signal processing, the computing efficiency of deep learning on DSPs is limited due to the lack of native hardware support. In this case, we present a contrary strategy and propose to enable signal processing on top of a classical deep learning accelerator (DLA). With the observation that irregular data patterns such as butterfly operations in FFT are the major barrier that hinders the deployment of signal processing on DLAs, we propose a programmable data shuffling fabric and have it inserted between the input buffer and computing array of DLAs such that the irregular data is reorganized and the processing is converted to be regular. With the online data shuffling, the proposed architecture, SigDLA, can adapt to various signal processing tasks without affecting the deep learning processing. Moreover, we build a reconfigurable computing array to suit the various data width requirements of both signal processing and deep learning. According to our experiments, SigDLA achieves an average performance speedup of 4.4$\times$, 1.4$\times$, and 1.52$\times$, and average energy reduction of 4.82$\times$, 3.27$\times$, and 2.15$\times$ compared to an embedded ARM processor with customized DSP instructions, a DSP processor, and an independent DSP-DLA architecture respectively with 17% more chip area over the original DLAs.

A typical optimization of customized accelerators for error-tolerant applications such as multimedia, recognition, and classification is to replace traditional arithmetic units like multipliers and adders with the approximate ones to enhance energy efficiency while adhering to accuracy requirements. However, the plethora of arithmetic units and diverse approximate unit options result in an exceedingly large design space. Therefore, there is a pressing need for an end-to-end design framework capable of navigating this intricate design space for approximation optimization. Traditional methods relying on simulation-based or blackbox model evaluations suffer from either high computational costs or limitations in accuracy and scalability, posing significant challenges to the optimization process. In this paper, we propose a Graph Neural Network (GNN) model that leverages the physical connections of arithmetic units to capture their influence on the performance, power, area (PPA), and accuracy of the accelerator. Particularly, we notice that critical path plays a key role in node feature of the GNN model and having it embedded in the feature vector greatly enhances the prediction quality of the models. On top of the models that allow rapid and efficient PPA and accuracy prediction of various approximate accelerator configurations, we can further explore the large design space effectively and build an end-to-end accelerator approximation framework named ApproxPilot to optimize the accelerator approximation. Our experimental results demonstrate that ApproxPilot outperforms state-of-the-art approximation optimization frameworks in both performance and hardware overhead with the same accuracy constraints.

Natural language interfaces have exhibited considerable potential in the automation of Verilog generation derived from high-level specifications through the utilization of large language models, garnering significant attention. Nevertheless, this paper elucidates that visual representations contribute essential contextual information critical to design intent for hardware architectures possessing spatial complexity, potentially surpassing the efficacy of natural-language-only inputs. Expanding upon this premise, our paper introduces an open-source benchmark for multi-modal generative models tailored for Verilog synthesis from visual-linguistic inputs, addressing both singular and complex modules. Additionally, we introduce an open-source visual and natural language Verilog query language framework to facilitate efficient and user-friendly multi-modal queries. To evaluate the performance of the proposed multi-modal hardware generative AI in Verilog generation tasks, we compare it with a popular method that relies solely on natural language. Our results demonstrate a significant accuracy improvement in the multi-modal generated Verilog compared to queries based solely on natural language. We hope to reveal a new approach to hardware design in the large-hardware-design-model era, thereby fostering a more diversified and productive approach to hardware design., Comment: Accepted by ICCAD 2024

Recent advances in large language models have demonstrated their potential for automated generation of hardware description language (HDL) code from high-level prompts. Researchers have utilized fine-tuning to enhance the ability of these large language models (LLMs) in the field of Chip Design. However, the lack of Verilog data hinders further improvement in the quality of Verilog generation by LLMs. Additionally, the absence of a Verilog and Electronic Design Automation (EDA) script data augmentation framework significantly increases the time required to prepare the training dataset for LLM trainers. This paper proposes an automated design-data augmentation framework, which generates high-volume and high-quality natural language aligned with Verilog and EDA scripts. For Verilog generation, it translates Verilog files to an abstract syntax tree and then maps nodes to natural language with a predefined template. For Verilog repair, it uses predefined rules to generate the wrong verilog file and then pairs EDA Tool feedback with the right and wrong verilog file. For EDA Script generation, it uses existing LLM(GPT-3.5) to obtain the description of the Script. To evaluate the effectiveness of our data augmentation method, we finetune Llama2-13B and Llama2-7B models using the dataset generated by our augmentation framework. The results demonstrate a significant improvement in the Verilog generation tasks with LLMs. Moreover, the accuracy of Verilog generation surpasses that of the current state-of-the-art open-source Verilog generation model, increasing from 58.8% to 70.6% with the same benchmark. Our 13B model (ChipGPT-FT) has a pass rate improvement compared with GPT-3.5 in Verilog generation and outperforms in EDA script (i.e., SiliconCompiler) generation with only 200 EDA script data., Comment: DAC 2024

Technology mapping is an essential step in EDA flow. However, the function of the circuit may be changed after technology mapping, and equivalence checking (EC) based verification is highly necessary. The traditional EC method has significant time and resource constraints, making it only feasible to carry out at a coarse-grained level. To make it efficient for technology mapping, we propose a fine-grained method called MEC, which leverages a combination of two approaches to significantly reduce the time cost of verification. The local block verification approach performs fast verification and the global graph cover approach guarantees correctness. The proposed method is rigorously tested and compared to three EC tools, and the results show that MEC technique offers a substantial improvement in speed. MEC not only offers a faster and more efficient way of performing EC on technology mapping but also opens up new opportunities for more fine-grained verification in the future., Comment: International Symposium of EDA 2023

Rewriting is a common procedure in logic synthesis aimed at improving the performance, power, and area (PPA) of circuits. The traditional reconvergence-driven And-Inverter Graph (AIG) rewriting method focuses solely on optimizing the reconvergence cone through Boolean algebra minimization. However, there exist opportunities to incorporate other node-rewriting algorithms that are better suited for specific cones. In this paper, we propose an adaptive reconvergence-driven AIG rewriting algorithm that combines two key techniques: multi-strategy-based AIG rewriting and strategy learning-based algorithm selection. The multi-strategy-based rewriting method expands upon the traditional approach by incorporating support for multi-node-rewriting algorithms, thus expanding the optimization space. Additionally, the strategy learning-based algorithm selection method determines the most suitable node-rewriting algorithm for a given cone. Experimental results demonstrate that our proposed method yields a significant average improvement of 5.567\% in size and 5.327\% in depth., Comment: The 41st IEEE International Conference on Computer Design

Fault-tolerant deep learning accelerator is the basis for highly reliable deep learning processing and critical to deploy deep learning in safety-critical applications such as avionics and robotics. Since deep learning is known to be computing- and memory-intensive, traditional fault-tolerant approaches based on redundant computing will incur substantial overhead including power consumption and chip area. To this end, we propose to characterize deep learning vulnerability difference across both neurons and bits of each neuron, and leverage the vulnerability difference to enable selective protection of the deep learning processing components from the perspective of architecture layer and circuit layer respectively. At the same time, we observe the correlation between model quantization and bit protection overhead of the underlying processing elements of deep learning accelerators, and propose to reduce the bit protection overhead by adding additional quantization constrain without compromising the model accuracy. Finally, we employ Bayesian optimization strategy to co-optimize the correlated cross-layer design parameters at algorithm layer, architecture layer, and circuit layer to minimize the hardware resource consumption while fulfilling multiple user constraints including reliability, accuracy, and performance of the deep learning processing at the same time., Comment: 16 pages, it has been presented at CCF-DAC 2023 while CCF-DAC does not own the copyright

While Federated learning (FL) is attractive for pulling privacy-preserving distributed training data, the credibility of participating clients and non-inspectable data pose new security threats, of which poisoning attacks are particularly rampant and hard to defend without compromising privacy, performance or other desirable properties of FL. To tackle this problem, we propose a self-purified FL (SPFL) method that enables benign clients to exploit trusted historical features of locally purified model to supervise the training of aggregated model in each iteration. The purification is performed by an attention-guided self-knowledge distillation where the teacher and student models are optimized locally for task loss, distillation loss and attention-based loss simultaneously. SPFL imposes no restriction on the communication protocol and aggregator at the server. It can work in tandem with any existing secure aggregation algorithms and protocols for augmented security and privacy guarantee. We experimentally demonstrate that SPFL outperforms state-of-the-art FL defenses against various poisoning attacks. The attack success rate of SPFL trained model is at most 3$\%$ above that of a clean model, even if the poisoning attack is launched in every iteration with all but one malicious clients in the system. Meantime, it improves the model quality on normal inputs compared to FedAvg, either under attack or in the absence of an attack.

Mixed-precision neural networks (MPNNs) that enable the use of just enough data width for a deep learning task promise significant advantages of both inference accuracy and computing overhead. FPGAs with fine-grained reconfiguration capability can adapt the processing with distinct data width and models, and hence, can theoretically unleash the potential of MPNNs. Nevertheless, commodity DPUs on FPGAs mostly emphasize generality and have limited support for MPNNs especially the ones with lower data width. In addition, primitive DSPs in FPGAs usually have much larger data width than that is required by MPNNs and haven't been sufficiently co-explored with MPNNs yet. To this end, we propose an open source MPNN accelerator design framework specifically tailored for FPGAs. In this framework, we have a systematic DSP-packing algorithm to pack multiple lower data width MACs in a single primitive DSP and enable efficient implementation of MPNNs. Meanwhile, we take DSP packing efficiency into consideration with MPNN quantization within a unified neural network architecture search (NAS) framework such that it can be aware of the DSP overhead during quantization and optimize the MPNN performance and accuracy concurrently. Finally, we have the optimized MPNN fine-tuned to a fully pipelined neural network accelerator template based on HLS and make best use of available resources for higher performance. Our experiments reveal the resulting accelerators produced by the proposed framework can achieve overwhelming advantages in terms of performance, resource utilization, and inference accuracy for MPNNs when compared with both handcrafted counterparts and prior hardware-aware neural network accelerators on FPGAs., Comment: Accepted by 2023 IEEE/ACM International Conference on Computer-Aided Design (ICCAD)

NPN classification has many applications in the synthesis and verification of digital circuits. The canonical-form-based method is the most common approach, designing a canonical form as representative for the NPN equivalence class first and then computing the transformation function according to the canonical form. Most works use variable symmetries and several signatures, mainly based on the cofactor, to simplify the canonical form construction and computation. This paper describes a novel canonical form and its computation algorithm by introducing Boolean influence to NPN classification, which is a basic concept in analysis of Boolean functions. We show that influence is input-negation-independent, input-permutation-dependent, and has other structural information than previous signatures for NPN classification. Therefore, it is a significant ingredient in speeding up NPN classification. Experimental results prove that influence plays an important role in reducing the transformation enumeration in computing the canonical form. Compared with the state-of-the-art algorithm implemented in ABC, our influence-aided canonical form for exact NPN classification gains up to 5.5x speedup., Comment: To be appeared in ICCAD'23

Winograd is generally utilized to optimize convolution performance and computational efficiency because of the reduced multiplication operations, but the reliability issues brought by winograd are usually overlooked. In this work, we observe the great potential of winograd convolution in improving neural network (NN) fault tolerance. Based on the observation, we evaluate winograd convolution fault tolerance comprehensively from different granularities ranging from models, layers, and operation types for the first time. Then, we explore the use of inherent fault tolerance of winograd convolution for cost-effective NN protection against soft errors. Specifically, we mainly investigate how winograd convolution can be effectively incorporated with classical fault-tolerant design approaches including triple modular redundancy (TMR), fault-aware retraining, and constrained activation functions. According to our experiments, winograd convolution can reduce the fault-tolerant design overhead by 55.77\% on average without any accuracy loss compared to standard convolution, and further reduce the computing overhead by 17.24\% when the inherent fault tolerance of winograd convolution is considered. When it is applied on fault-tolerant neural networks enhanced with fault-aware retraining and constrained activation functions, the resulting model accuracy generally shows significant improvement in presence of various faults.

Open-source EDA shows promising potential in unleashing EDA innovation and lowering the cost of chip design. This paper presents an open-source EDA project, iEDA, aiming for building a basic infrastructure for EDA technology evolution and closing the industrial-academic gap in the EDA area. iEDA now covers the whole flow of physical design (including Floorplan, Placement, CTS, Routing, Timing Optimization etc.), and part of the analysis tools (Static Timing Analysis and Power Analysis). To demonstrate the effectiveness of iEDA, we implement and tape out three chips of different scales (from 700k to 1.5M gates) on different process nodes (110nm and 28nm) with iEDA. iEDA is publicly available from the project home page http://ieda.oscc.cc.

To ensure resilient neural network processing on even unreliable hardware, comprehensive reliability analysis against various hardware faults is generally required before the deep neural network models are deployed, and efficient error injection tools are highly demanded. However, most existing fault injection tools remain rather limited to basic fault injection to neurons and fail to provide fine-grained vulnerability analysis capability. In addition, many of the fault injection tools still need to change the neural network models and make the fault injection closely coupled with normal neural network processing, which further complicates the use of the fault injection tools and slows down the fault simulation. In this work, we propose MRFI, a highly configurable multi-resolution fault injection tool for deep neural networks. It enables users to modify an independent fault configuration file rather than neural network models for the fault injection and vulnerability analysis. Particularly, it integrates extensive fault analysis functionalities from different perspectives and enables multi-resolution investigation of the vulnerability of neural networks. In addition, it does not modify the major neural network computing framework of PyTorch. Hence, it allows parallel processing on GPUs naturally and exhibits fast fault simulation according to our experiments., Comment: 9 pages, 12 figures, source code is on https://github.com/fffasttime/MRFI

As large language models (LLMs) like ChatGPT exhibited unprecedented machine intelligence, it also shows great performance in assisting hardware engineers to realize higher-efficiency logic design via natural language interaction. To estimate the potential of the hardware design process assisted by LLMs, this work attempts to demonstrate an automated design environment that explores LLMs to generate hardware logic designs from natural language specifications. To realize a more accessible and efficient chip development flow, we present a scalable four-stage zero-code logic design framework based on LLMs without retraining or finetuning. At first, the demo, ChipGPT, begins by generating prompts for the LLM, which then produces initial Verilog programs. Second, an output manager corrects and optimizes these programs before collecting them into the final design space. Eventually, ChipGPT will search through this space to select the optimal design under the target metrics. The evaluation sheds some light on whether LLMs can generate correct and complete hardware logic designs described by natural language for some specifications. It is shown that ChipGPT improves programmability, and controllability, and shows broader design optimization space compared to prior work and native LLMs alone.

Vision Transformers (ViTs) that leverage self-attention mechanism have shown superior performance on many classical vision tasks compared to convolutional neural networks (CNNs) and gain increasing popularity recently. Existing ViTs works mainly optimize performance and accuracy, but ViTs reliability issues induced by soft errors in large-scale VLSI designs have generally been overlooked. In this work, we mainly study the reliability of ViTs and investigate the vulnerability from different architecture granularities ranging from models, layers, modules, and patches for the first time. The investigation reveals that ViTs with the self-attention mechanism are generally more resilient on linear computing including general matrix-matrix multiplication (GEMM) and full connection (FC) and show a relatively even vulnerability distribution across the patches. ViTs involve more fragile non-linear computing such as softmax and GELU compared to typical CNNs. With the above observations, we propose a lightweight block-wise algorithm-based fault tolerance (LB-ABFT) approach to protect the linear computing implemented with distinct sizes of GEMM and apply a range-based protection scheme to mitigate soft errors in non-linear computing. According to our experiments, the proposed fault-tolerant approaches enhance ViTs accuracy significantly with minor computing overhead in presence of various soft errors.

Vision Transformers (ViTs) with outstanding performance becomes a popular backbone of deep learning models for the main-stream vision tasks including classification, object detection, and segmentation. Other than the performance, reliability is also a critical metric for the adoption of ViTs in safety-critical applications such as autonomous driving and robotics. With the observation that the major computing blocks in ViTs such as multi-head attention and feed forward are usually performed with general matrix multiplication (GEMM), we propose a classical algorithm-based fault tolerance (ABFT) strategy originally developed for GEMM to protect ViTs against soft errors in the underlying computing engines. Unlike classical ABFT that will invoke the expensive error recovery procedure whenever computing errors are detected, we leverage the inherent fault-tolerance of ViTs and propose an approximate ABFT, namely ApproxABFT, to invoke the error recovery procedure only when the computing errors are significant enough, which skips many useless error recovery procedures and simplifies the overall GEMM error recovery. Meanwhile, it also relaxes the error threshold in error recovery procedure and ignores minor computing errors, which reduces the error recovery complexity and improves the error recovery quality. In addition, we also apply a fine-grained blocking strategy to ApproxABFT and split GEMM with distinct sizes into smaller sub blocks such that it can smooth the error thresholds across ViTs and further improve the error recovery quality. According to our experiments, the ApproxABFT reduces the computing overhead by 25.92\% to 81.62\% and improves the model accuracy by 2.63\% to 72.56\% compared to the baseline ABFT while the blocking optimization further reduces the computing overhead by 6.56\% to 73.5\% with comparable accuracy.

NPN classification is an essential problem in the design and verification of digital circuits. Most existing works explored variable symmetries and cofactor signatures to develop their classification methods. However, cofactor signatures only consider the face characteristics of Boolean functions. In this paper, we propose a new NPN classifier using both face and point characteristics of Boolean functions, including cofactor, influence, and sensitivity. The new method brings a new perspective to the classification of Boolean functions. The classifier only needs to compute some signatures, and the equality of corresponding signatures is a prerequisite for NPN equivalence. Therefore, these signatures can be directly used for NPN classification, thus avoiding the exhaustive transformation enumeration. The experiments show that the proposed NPN classifier gains better NPN classification accuracy with comparable speed., Comment: Accepted to DATE'23

Soft errors in large VLSI circuits pose dramatic influence on computing- and memory-intensive neural network (NN) processing. Understanding the influence of soft errors on NNs is critical to protect against soft errors for reliable NN processing. Prior work mainly rely on fault simulation to analyze the influence of soft errors on NN processing. They are accurate but usually specific to limited configurations of errors and NN models due to the prohibitively slow simulation speed especially for large NN models and datasets. With the observation that the influence of soft errors propagates across a large number of neurons and accumulates as well, we propose to characterize the soft error induced data disturbance on each neuron with normal distribution model according to central limit theorem and develop a series of statistical models to analyze the behavior of NN models under soft errors in general. The statistical models reveal not only the correlation between soft errors and NN model accuracy, but also how NN parameters such as quantization and architecture affect the reliability of NNs. The proposed models are compared with fault simulation and verified comprehensively. In addition, we observe that the statistical models that characterize the soft error influence can also be utilized to predict fault simulation results in many cases and we explore the use of the proposed statistical models to accelerate fault simulations of NNs. According to our experiments, the accelerated fault simulation shows almost two orders of magnitude speedup with negligible simulation accuracy loss over the baseline fault simulations.

Secure aggregation is widely used in horizontal Federated Learning (FL), to prevent leakage of training data when model updates from data owners are aggregated. Secure aggregation protocols based on Homomorphic Encryption (HE) have been utilized in industrial cross-silo FL systems, one of the settings involved with privacy-sensitive organizations such as financial or medical, presenting more stringent requirements on privacy security. However, existing HE-based solutions have limitations in efficiency and security guarantees against colluding adversaries without a Trust Third Party. This paper proposes an efficient Doubly Homomorphic Secure Aggregation (DHSA) scheme for cross-silo FL, which utilizes multi-key Homomorphic Encryption (MKHE) and seed homomorphic pseudorandom generator (SHPRG) as cryptographic primitives. The application of MKHE provides strong security guarantees against up to $N-2$ participates colluding with the aggregator, with no TTP required. To mitigate the large computation and communication cost of MKHE, we leverage the homomorphic property of SHPRG to replace the majority of MKHE computation by computationally-friendly mask generation from SHPRG, while preserving the security. Overall, the resulting scheme satisfies the stringent security requirements of typical cross-silo FL scenarios, at the same time providing high computation and communication efficiency for practical usage. We experimentally demonstrate our scheme brings a speedup to 20$\times$ over the state-of-the-art HE-based secure aggregation, and reduces the traffic volume to approximately 1.5$\times$ inflation over the plain learning setting., Comment: 34 pages, 6 figures

With the rapid advancements of deep learning in the past decade, it can be foreseen that deep learning will be continuously deployed in more and more safety-critical applications such as autonomous driving and robotics. In this context, reliability turns out to be critical to the deployment of deep learning in these applications and gradually becomes a first-class citizen among the major design metrics like performance and energy efficiency. Nevertheless, the back-box deep learning models combined with the diverse underlying hardware faults make resilient deep learning extremely challenging. In this special session, we conduct a comprehensive survey of fault-tolerant deep learning design approaches with a hierarchical perspective and investigate these approaches from model layer, architecture layer, circuit layer, and cross layer respectively., Comment: Special session submitted to VTS'22

To prevent private training data leakage in Fed?erated Learning systems, we propose a novel se?cure aggregation scheme based on seed homomor?phic pseudo-random generator (SHPRG), named SASH. SASH leverages the homomorphic property of SHPRG to simplify the masking and demask?ing scheme, which for each of the clients and for the server, entails an overhead linear w.r.t model size and constant w.r.t number of clients. We prove that even against worst-case colluding adversaries, SASH preserves training data privacy, while being resilient to dropouts without extra overhead. We experimentally demonstrate SASH significantly improves the efficiency to 20 times over baseline, especially in the more realistic case where the numbers of clients and model size become large, and a cer?tain percentage of clients drop out from the system., Comment: 6 pages, 3 figures. The 38th Conference on Uncertainty in Artificial Intelligence. 2022

Carbon nanotube field-effect transistors (CNFET) emerge as a promising alternative to CMOS transistors for the much higher speed and energy efficiency, which makes the technology particularly suitable for building the energy-hungry last level cache (LLC). However, the process variations (PVs) in CNFET caused by the imperfect fabrication lead to large timing variation and the worst-case timing dramatically limits the LLC operation speed. Particularly, we observe that the CNFET-based cache latency distribution is closely related to the LLC layouts. For the two typical LLC layouts that have the CNT growth direction aligned to the cache way direction and cache set direction respectively, we proposed variation-aware set aligned (VASA) cache and variation-aware way aligned (VAWA) cache in combination with corresponding cache optimizations such as data shuffling and page mapping to enable low-latency cache for frequently used data. According to our experiments, the optimized LLC reduces the average access latency by 32% and 45% compared to the baseline designs on the two different CNFET layouts respectively while it improves the overall performance by 6\% and 9\% and reduces the energy consumption by 4% and 8% respectively. In addition, with both the architecture induced latency variation and PV incurred latency variation considered in a unified model, we extended the VAWA and VASA cache design for the CNFET-based NUCA and the proposed NUCA achieves both significant performance improvement and energy saving compared to the straightforward variation-aware NUCA., Comment: 13 pages, extended on top of "Exploring emerging CNFET for efficient last level cache design" published in ASPDAC'19

AIoT processors fabricated with newer technology nodes suffer rising soft errors due to the shrinking transistor sizes and lower power supply. Soft errors on the AIoT processors particularly the deep learning accelerators (DLAs) with massive computing may cause substantial computing errors. These computing errors are difficult to be captured by the conventional training on general purposed processors like CPUs and GPUs in a server. Applying the offline trained neural network models to the edge accelerators with errors directly may lead to considerable prediction accuracy loss. To address the problem, we propose a remote retraining framework (R2F) for remote AIoT processors with computing errors. It takes the remote AIoT processor with soft errors in the training loop such that the on-site computing errors can be learned with the application data on the server and the retrained models can be resilient to the soft errors. Meanwhile, we propose an optimized partial TMR strategy to enhance the retraining. According to our experiments, R2F enables elastic design trade-offs between the model accuracy and the performance penalty. The top-5 model accuracy can be improved by 1.93%-13.73% with 0%-200% performance penalty at high fault error rate. In addition, we notice that the retraining requires massive data transmission and even dominates the training time, and propose a sparse increment compression approach for the data transmission optimization, which reduces the retraining time by 38%-88% on average with negligible accuracy loss over a straightforward remote retraining.

Deformable convolution networks (DCNs) proposed to address the image recognition with geometric or photometric variations typically involve deformable convolution that convolves on arbitrary locations of input features. The locations change with different inputs and induce considerable dynamic and irregular memory accesses which cannot be handled by classic neural network accelerators (NNAs). Moreover, bilinear interpolation (BLI) operation that is required to obtain deformed features in DCNs also cannot be deployed on existing NNAs directly. Although a general purposed processor (GPP) seated along with classic NNAs can process the deformable convolution, the processing on GPP can be extremely slow due to the lack of parallel computing capability. To address the problem, we develop a DCN accelerator on existing NNAs to support both the standard convolution and deformable convolution. Specifically, for the dynamic and irregular accesses in DCNs, we have both the input and output features divided into tiles and build a tile dependency table (TDT) to track the irregular tile dependency at runtime. With the TDT, we further develop an on-chip tile scheduler to handle the dynamic and irregular accesses efficiently. In addition, we propose a novel mapping strategy to enable parallel BLI processing on NNAs and apply layer fusion techniques for more energy-efficient DCN processing. According to our experiments, the proposed accelerator achieves orders of magnitude higher performance and energy efficiency compared to the typical computing architectures including ARM, ARM+TPU, and GPU with 6.6\% chip area penalty to a classic NNA.

Hardware faults on the regular 2-D computing array of a typical deep learning accelerator (DLA) can lead to dramatic prediction accuracy loss. Prior redundancy design approaches typically have each homogeneous redundant processing element (PE) to mitigate faulty PEs for a limited region of the 2-D computing array rather than the entire computing array to avoid the excessive hardware overhead. However, they fail to recover the computing array when the number of faulty PEs in any region exceeds the number of redundant PEs in the same region. The mismatch problem deteriorates when the fault injection rate rises and the faults are unevenly distributed. To address the problem, we propose a hybrid computing architecture (HyCA) for fault-tolerant DLAs. It has a set of dot-production processing units (DPPUs) to recompute all the operations that are mapped to the faulty PEs despite the faulty PE locations. According to our experiments, HyCA shows significantly higher reliability, scalability, and performance with less chip area penalty when compared to the conventional redundancy approaches. Moreover, by taking advantage of the flexible recomputing, HyCA can also be utilized to scan the entire 2-D computing array and detect the faulty PEs effectively at runtime.