Search

While large language models (LLMs) have demonstrated the ability to generate hardware description language (HDL) code for digital circuits, they still suffer from the hallucination problem, which leads to the generation of incorrect HDL code or misunderstanding of specifications. In this work, we introduce a human-expert-inspired method to mitigate the hallucination of LLMs and improve the performance in HDL code generation. We first let LLMs classify the type of the circuit based on the specifications. Then, according to the type of the circuit, we split the tasks into several sub-procedures, including information extraction and human-like design flow using Electronic Design Automation (EDA) tools. Besides, we also use a search method to mitigate the variation in code generation. Experimental results show that our method can significantly improve the functional correctness of the generated Verilog and reduce the hallucination of LLMs.

In digital circuit design, testbenches constitute the cornerstone of simulation-based hardware verification. Traditional methodologies for testbench generation during simulation-based hardware verification still remain partially manual, resulting in inefficiencies in testing various scenarios and requiring expensive time from designers. Large Language Models (LLMs) have demonstrated their potential in automating the circuit design flow. However, directly applying LLMs to generate testbenches suffers from a low pass rate. To address this challenge, we introduce AutoBench, the first LLM-based testbench generator for digital circuit design, which requires only the description of the design under test (DUT) to automatically generate comprehensive testbenches. In AutoBench, a hybrid testbench structure and a self-checking system are realized using LLMs. To validate the generated testbenches, we also introduce an automated testbench evaluation framework to evaluate the quality of generated testbenches from multiple perspectives. Experimental results demonstrate that AutoBench achieves a 57% improvement in the testbench pass@1 ratio compared with the baseline that directly generates testbenches using LLMs. For 75 sequential circuits, AutoBench successfully has a 3.36 times testbench pass@1 ratio compared with the baseline. The source codes and experimental results are open-sourced at this link: https://github.com/AutoBench/AutoBench

In High-Level Synthesis (HLS), converting a regular C/C++ program into its HLS-compatible counterpart (HLS-C) still requires tremendous manual effort. Various program scripts have been introduced to automate this process. But the resulting codes usually contain many issues that should be manually repaired by developers. Since Large Language Models (LLMs) have the ability to automate code generation, they can also be used for automated program repair in HLS. However, due to the limited training of LLMs considering hardware and software simultaneously, hallucinations may occur during program repair using LLMs, leading to compilation failures. Besides, using LLMs for iterative repair also incurs a high cost. To address these challenges, we propose an LLM-driven program repair framework that takes regular C/C++ code as input and automatically generates its corresponding HLS-C code for synthesis while minimizing human repair effort. To mitigate the hallucinations in LLMs and enhance the prompt quality, a Retrieval-Augmented Generation (RAG) paradigm is introduced to guide the LLMs toward correct repair. In addition, we use LLMs to create a static bit width optimization program to identify the optimized bit widths for variables. Moreover, LLM-driven HLS optimization strategies are introduced to add/tune pragmas in HLS-C programs for circuit optimization. Experimental results demonstrate that the proposed LLM-driven automated framework can achieve much higher repair pass rates in 24 real-world applications compared with the traditional scripts and the direct application of LLMs for program repair.

Deep neural networks (DNNs) have made breakthroughs in various fields including image recognition and language processing. DNNs execute hundreds of millions of multiply-and-accumulate (MAC) operations. To efficiently accelerate such computations, analog in-memory-computing platforms have emerged leveraging emerging devices such as resistive RAM (RRAM). However, such accelerators face the hurdle of being required to have sufficient on-chip crossbars to hold all the weights of a DNN. Otherwise, RRAM cells in the crossbars need to be reprogramed to process further layers, which causes huge time/energy overhead due to the extremely slow writing and verification of the RRAM cells. As a result, it is still not possible to deploy such accelerators to process large-scale DNNs in industry. To address this problem, we propose the BasisN framework to accelerate DNNs on any number of available crossbars without reprogramming. BasisN introduces a novel representation of the kernels in DNN layers as combinations of global basis vectors shared between all layers with quantized coefficients. These basis vectors are written to crossbars only once and used for the computations of all layers with marginal hardware modification. BasisN also provides a novel training approach to enhance computation parallelization with the global basis vectors and optimize the coefficients to construct the kernels. Experimental results demonstrate that cycles per inference and energy-delay product were reduced to below 1% compared with applying reprogramming on crossbars in processing large-scale DNNs such as DenseNet and ResNet on ImageNet and CIFAR100 datasets, while the training and hardware costs are negligible., Comment: accepted by ICCAD2024

In this paper, we introduce a novel low-latency inference framework for large language models (LLMs) inference which enables LLMs to perform inferences with incomplete prompts. By reallocating computational processes to prompt input phase, we achieve a substantial reduction in latency, thereby significantly enhancing the interactive experience for users of LLMs. The framework adeptly manages the visibility of the streaming prompt to the model, allowing it to infer from incomplete prompts or await additional prompts. Compared with traditional inference methods that utilize complete prompts, our approach demonstrates an average reduction of 59% in response latency on the MMLU-Pro dataset, while maintaining comparable accuracy. Additionally, our framework facilitates collaborative inference and output across different models. By employing an LLM for inference and a small language model (SLM) for output, we achieve an average 68% reduction in response latency, alongside a 5.5% improvement in accuracy on the MMLU-Pro dataset compared with the SLM baseline. For long prompts exceeding 20 sentences, the response latency can be reduced by up to 93%.

Deep neural networks (DNNs) have achieved great breakthroughs in many fields such as image classification and natural language processing. However, the execution of DNNs needs to conduct massive numbers of multiply-accumulate (MAC) operations on hardware and thus incurs a large power consumption. To address this challenge, we propose a novel digital MAC design based on encoding. In this new design, the multipliers are replaced by simple logic gates to project the results onto a wide bit representation. These bits carry individual position weights, which can be trained for specific neural networks to enhance inference accuracy. The outputs of the new multipliers are added by bit-wise weighted accumulation and the accumulation results are compatible with existing computing platforms accelerating neural networks with either uniform or non-uniform quantization. Since the multiplication function is replaced by simple logic projection, the critical paths in the resulting circuits become much shorter. Correspondingly, pipelining stages in the MAC array can be reduced, leading to a significantly smaller area as well as a better power efficiency. The proposed design has been synthesized and verified by ResNet18-Cifar10, ResNet20-Cifar100 and ResNet50-ImageNet. The experimental results confirmed the reduction of circuit area by up to 79.63% and the reduction of power consumption of executing DNNs by up to 70.18%, while the accuracy of the neural networks can still be well maintained.

Deep neural networks (DNNs) have demonstrated remarkable success in various fields. However, the large number of floating-point operations (FLOPs) in DNNs poses challenges for their deployment in resource-constrained applications, e.g., edge devices. To address the problem, pruning has been introduced to reduce the computational cost in executing DNNs. Previous pruning strategies are based on weight values, gradient values and activation outputs. Different from previous pruning solutions, in this paper, we propose a class-aware pruning technique to compress DNNs, which provides a novel perspective to reduce the computational cost of DNNs. In each iteration, the neural network training is modified to facilitate the class-aware pruning. Afterwards, the importance of filters with respect to the number of classes is evaluated. The filters that are only important for a few number of classes are removed. The neural network is then retrained to compensate for the incurred accuracy loss. The pruning iterations end until no filter can be removed anymore, indicating that the remaining filters are very important for many classes. This pruning technique outperforms previous pruning solutions in terms of accuracy, pruning ratio and the reduction of FLOPs. Experimental results confirm that this class-aware pruning technique can significantly reduce the number of weights and FLOPs, while maintaining a high inference accuracy., Comment: Accepted by Design Automation and Test in Europe (DATE) 2024

Having the potential for high speed, high throughput, and low energy cost, optical neural networks (ONNs) have emerged as a promising candidate for accelerating deep learning tasks. In conventional ONNs, light amplitudes are modulated at the input and detected at the output. However, the light phases are still ignored in conventional structures, although they can also carry information for computing. To address this issue, in this paper, we propose a framework called OplixNet to compress the areas of ONNs by modulating input image data into the amplitudes and phase parts of light signals. The input and output parts of the ONNs are redesigned to make full use of both amplitude and phase information. Moreover, mutual learning across different ONN structures is introduced to maintain the accuracy. Experimental results demonstrate that the proposed framework significantly reduces the areas of ONNs with the accuracy within an acceptable range. For instance, 75.03% area is reduced with a 0.33% accuracy decrease on fully connected neural network (FCNN) and 74.88% area is reduced with a 2.38% accuracy decrease on ResNet-32., Comment: Accepted by Design Automation and Test in Europe (DATE) 2024

Deep neural networks (DNNs) have been successfully applied in various fields. In DNNs, a large number of multiply-accumulate (MAC) operations are required to be performed, posing critical challenges in applying them in resource-constrained platforms, e.g., edge devices. To address this challenge, in this paper, we propose a class-based early-exit for dynamic inference. Instead of pushing DNNs to make a dynamic decision at intermediate layers, we take advantage of the learned features in these layers to exclude as many irrelevant classes as possible, so that later layers only have to determine the target class among the remaining classes. When only one class remains at a layer, this class is the corresponding classification result. Experimental results demonstrate the computational cost of DNNs in inference can be reduced significantly with the proposed early-exit technique. The codes can be found at https://github.com/HWAI-TUDa/EarlyClassExclusion.

Neural networks (NNs) have been successfully deployed in various fields. In NNs, a large number of multiplyaccumulate (MAC) operations need to be performed. Most existing digital hardware platforms rely on parallel MAC units to accelerate these MAC operations. However, under a given area constraint, the number of MAC units in such platforms is limited, so MAC units have to be reused to perform MAC operations in a neural network. Accordingly, the throughput in generating classification results is not high, which prevents the application of traditional hardware platforms in extreme-throughput scenarios. Besides, the power consumption of such platforms is also high, mainly due to data movement. To overcome this challenge, in this paper, we propose to flatten and implement all the operations at neurons, e.g., MAC and ReLU, in a neural network with their corresponding logic circuits. To improve the throughput and reduce the power consumption of such logic designs, the weight values are embedded into the MAC units to simplify the logic, which can reduce the delay of the MAC units and the power consumption incurred by weight movement. The retiming technique is further used to improve the throughput of the logic circuits for neural networks. In addition, we propose a hardware-aware training method to reduce the area of logic designs of neural networks. Experimental results demonstrate that the proposed logic designs can achieve high throughput and low power consumption for several high-throughput applications., Comment: accepted by ASPDAC 2024

Convolutional neural networks (CNNs) have been successfully applied in a range of fields such as image classification and object segmentation. To improve their expressivity, various techniques, such as novel CNN architectures, have been explored. However, the performance gain from such techniques tends to diminish. To address this challenge, many researchers have shifted their focus to increasing the non-linearity of neurons, the fundamental building blocks of neural networks, to enhance the network expressivity. Nevertheless, most of these approaches incur a large number of parameters and thus formidable computation cost inevitably, impairing their efficiency to be deployed in practice. In this work, an efficient quadratic neuron structure is proposed to preserve the non-linearity with only negligible parameter and computation cost overhead. The proposed quadratic neuron can maximize the utilization of second-order computation information to improve the network performance. The experimental results have demonstrated that the proposed quadratic neuron can achieve a higher accuracy and a better computation efficiency in classification tasks compared with both linear neurons and non-linear neurons from previous works.

Deep neural networks (DNNs) have been successfully applied in various fields. A major challenge of deploying DNNs, especially on edge devices, is power consumption, due to the large number of multiply-and-accumulate (MAC) operations. To address this challenge, we propose PowerPruning, a novel method to reduce power consumption in digital neural network accelerators by selecting weights that lead to less power consumption in MAC operations. In addition, the timing characteristics of the selected weights together with all activation transitions are evaluated. The weights and activations that lead to small delays are further selected. Consequently, the maximum delay of the sensitized circuit paths in the MAC units is reduced even without modifying MAC units, which thus allows a flexible scaling of supply voltage to reduce power consumption further. Together with retraining, the proposed method can reduce power consumption of DNNs on hardware by up to 78.3% with only a slight accuracy loss.

In deep neural networks (DNNs), there are a huge number of weights and multiply-and-accumulate (MAC) operations. Accordingly, it is challenging to apply DNNs on resource-constrained platforms, e.g., mobile phones. Quantization is a method to reduce the size and the computational complexity of DNNs. Existing quantization methods either require hardware overhead to achieve a non-uniform quantization or focus on model-wise and layer-wise uniform quantization, which are not as fine-grained as filter-wise quantization. In this paper, we propose a class-based quantization method to determine the minimum number of quantization bits for each filter or neuron in DNNs individually. In the proposed method, the importance score of each filter or neuron with respect to the number of classes in the dataset is first evaluated. The larger the score is, the more important the filter or neuron is and thus the larger the number of quantization bits should be. Afterwards, a search algorithm is adopted to exploit the different importance of filters and neurons to determine the number of quantization bits of each filter or neuron. Experimental results demonstrate that the proposed method can maintain the inference accuracy with low bit-width quantization. Given the same number of quantization bits, the proposed method can also achieve a better inference accuracy than the existing methods., Comment: accepted by DATE2023 (Design, Automation and Test in Europe)

Deep neural networks (DNNs) have successfully been applied in many fields in the past decades. However, the increasing number of multiply-and-accumulate (MAC) operations in DNNs prevents their application in resource-constrained and resource-varying platforms, e.g., mobile phones and autonomous vehicles. In such platforms, neural networks need to provide acceptable results quickly and the accuracy of the results should be able to be enhanced dynamically according to the computational resources available in the computing system. To address these challenges, we propose a design framework called SteppingNet. SteppingNet constructs a series of subnets whose accuracy is incrementally enhanced as more MAC operations become available. Therefore, this design allows a trade-off between accuracy and latency. In addition, the larger subnets in SteppingNet are built upon smaller subnets, so that the results of the latter can directly be reused in the former without recomputation. This property allows SteppingNet to decide on-the-fly whether to enhance the inference accuracy by executing further MAC operations. Experimental results demonstrate that SteppingNet provides an effective incremental accuracy improvement and its inference accuracy consistently outperforms the state-of-the-art work under the same limit of computational resources., Comment: accepted by DATE2023 (Design, Automation and Test in Europe)

The last decade has witnessed the breakthrough of deep neural networks (DNNs) in many fields. With the increasing depth of DNNs, hundreds of millions of multiply-and-accumulate (MAC) operations need to be executed. To accelerate such operations efficiently, analog in-memory computing platforms based on emerging devices, e.g., resistive RAM (RRAM), have been introduced. These acceleration platforms rely on analog properties of the devices and thus suffer from process variations and noise. Consequently, weights in neural networks configured into these platforms can deviate from the expected values, which may lead to feature errors and a significant degradation of inference accuracy. To address this issue, in this paper, we propose a framework to enhance the robustness of neural networks under variations and noise. First, a modified Lipschitz constant regularization is proposed during neural network training to suppress the amplification of errors propagated through network layers. Afterwards, error compensation is introduced at necessary locations determined by reinforcement learning to rescue the feature maps with remaining errors. Experimental results demonstrate that inference accuracy of neural networks can be recovered from as low as 1.69% under variations and noise back to more than 95% of their original accuracy, while the training and hardware cost are negligible., Comment: Accepted by DATE 2023 (Design, Automation and Test in Europe)

In digital circuit designs, sequential components such as flip-flops are used to synchronize signal propagations. Logic computations are aligned at and thus isolated by flip-flop stages. Although this fully synchronous style can reduce design efforts significantly, it may affect circuit performance negatively, because sequential components can only introduce delays into signal propagations but never accelerate them. In this paper, we propose a new timing model, VirtualSync+, in which signals, specially those along critical paths, are allowed to propagate through several sequential stages without flip-flops. Timing constraints are still satisfied at the boundary of the optimized circuit to maintain a consistent interface with existing designs. By removing clock-to-q delays and setup time requirements of flip-flops on critical paths, the performance of a circuit can be pushed even beyond the limit of traditional sequential designs. In addition, we further enhance the optimization with VirtualSync+ by fine-tuning with commercial design tools, e.g., Design Compiler from Synopsys, to achieve more accurate result. Experimental results demonstrate that circuit performance can be improved by up to 4% (average 1.5%) compared with that after extreme retiming and sizing, while the increase of area is still negligible. This timing performance is enhanced beyond the limit of traditional sequential designs. It also demonstrates that compared with those after retiming and sizing, the circuits with VirtualSync+ can achieve better timing performance under the same area cost or smaller area cost under the same clock period, respectively.

With recent advances in reverse engineering, attackers can reconstruct a netlist to counterfeit chips by opening the die and scanning all layers of authentic chips. This relatively easy counterfeiting is made possible by the use of the standard simple clocking scheme, where all combinational blocks function within one clock period, so that a netlist of combinational logic gates and flip-flops is sufficient to duplicate a design. In this paper, we propose to invalidate the assumption that a netlist completely represents the function of a circuit with unconventional timing. With the introduced wave-pipelining paths, attackers have to capture gate and interconnect delays during reverse engineering, or to test a huge number of combinational paths to identify the wave-pipelining paths. To hinder the test-based attack, we construct false paths with wave-pipelining to increase the counterfeiting challenge. Experimental results confirm that wave-pipelining true paths and false paths can be constructed in benchmark circuits successfully with only a negligible cost, thus thwarting the potential attack techniques.

In static timing analysis, clock-to-q delays of flip-flops are considered as constants. Setup times and hold times are characterized separately and also used as constants. The characterized delays, setup times and hold times, are ap- plied in timing analysis independently to verify the perfor- mance of circuits. In reality, however, clock-to-q delays of flip-flops depend on both setup and hold times. Instead of being constants, these delays change with respect to different setup/hold time combinations. Consequently, the simple ab- straction of setup/hold times and constant clock-to-q delays introduces inaccuracy in timing analysis. In this paper, we propose a holistic method to consider the relation between clock-to-q delays and setup/hold time combinations with a piecewise linear model. The result is more accurate than that of traditional timing analysis, and the incorporation of the interdependency between clock-to-q delays, setup times and hold times may also improve circuit performance., Comment: IEEE/ACM International Conference on Computer-Aided Design (ICCAD), November 2016

At nanometer manufacturing technology nodes, process variations significantly affect circuit performance. To combat them, post- silicon clock tuning buffers can be deployed to balance timing bud- gets of critical paths for each individual chip after manufacturing. The challenge of this method is that path delays should be mea- sured for each chip to configure the tuning buffers properly. Current methods for this delay measurement rely on path-wise frequency stepping. This strategy, however, requires too much time from ex- pensive testers. In this paper, we propose an efficient delay test framework (EffiTest) to solve the post-silicon testing problem by aligning path delays using the already-existing tuning buffers in the circuit. In addition, we only test representative paths and the delays of other paths are estimated by statistical delay prediction. Exper- imental results demonstrate that the proposed method can reduce the number of frequency stepping iterations by more than 94% with only a slight yield loss., Comment: ACM/IEEE Design Automation Conference (DAC), June 2016

At submicron manufacturing technology nodes process variations affect circuit performance significantly. This trend leads to a large timing margin and thus overdesign to maintain yield. To combat this pessimism, post-silicon clock tuning buffers can be inserted into circuits to balance timing budgets of critical paths with their neighbors. After manufacturing, these clock buffers can be configured for each chip individually so that chips with timing failures may be rescued to improve yield. In this paper, we propose a sampling-based method to determine the proper locations of these buffers. The goal of this buffer insertion is to reduce the number of buffers and their ranges, while still maintaining a good yield improvement. Experimental results demonstrate that our algorithm can achieve a significant yield improvement (up to 35%) with only a small number of buffers., Comment: Design, Automation and Test in Europe (DATE), 2016