Detailed routing constitutes a critical phase in the very large-scale integration (VLSI) physical design, widely regarded as the most time-consuming and computationally intensive step in the back-end design process. Due to its iterative nature and strong data dependencies, conventional parallel acceleration techniques often suffer from limited scalability and effectiveness. To address these challenges, we propose RouterAcc, an FPGA-based software–hardware co-design acceleration framework tailored for VLSI detailed routing. RouterAcc incorporates an access analysis mechanism and a termination condition strategy to accelerate convergence. Furthermore, we employ a hierarchical storage mapping scheme and a flexible dimension-partitioning architecture to alleviate memory bottlenecks and enhance data locality. Additionally, RouterAcc leverages a hierarchical comparison pipeline with fully parallelized computing units and a data preprocessing strategy to maximize computational efficiency. Experimental results on the ISPD’18 benchmarks demonstrate that RouterAcc achieves consistent speedups of 2.1×–2.3× over TritonRoute with less than 1% quality degradation. With further co-optimization, RouterAcc attains speedups of 2.7×–11.8× while maintaining routing quality comparable to TritonRoute and surpassing Dr.CU 2.0 as well as the state-of-the-art (SOTA) FPGA-based approaches.

Hyperdimensional computing (HDC) has emerged as a promising cognitive computing paradigm, offering exceptional robustness and energy efficiency for intelligent applications. However, the computational demands of HDC, particularly during the encoding and associative search phases, pose significant challenges due to their time and resource intensity. In this paper, we propose LUT-HD, a software-hardware co-design framework that accelerates HDC inference by leveraging efficient table lookup techniques. First, we introduce a binary code quantization (BCQ) algorithm based on a lookup table (LUT) that transforms costly matrix-vector multiplications in HDC into simple table lookups using precomputed results. Next, we propose a custom FPGA-based accelerator tailored for LUT-based HDC to strike a balance between accuracy and efficiency. This accelerator incorporates a performance-optimized pipeline for encoding and associative search, enhancing computational speed and resource utilization. Experimental results demonstrate that LUT-HD achieves up to 14.6 × inference speedup and reduces 97.3% energy consumption compared to the GPU platform. In addition, compared to state-of-the-art (SOTA) HDC solutions, LUT-HD offers a 5.5× speedup with negligible accuracy loss and reduces 44.8% energy consumption.

The click-through rate (CTR) prediction task is to predict whether a user will click on the recommended item. As mind-boggling amounts of data are produced online daily, accelerating CTR prediction model training is critical to ensuring an up-to-date model and reducing the training cost. One approach to increase the training speed is to apply large batch training. However, as shown in computer vision and natural language processing tasks, training with a large batch easily suffers from the loss of accuracy. Our experiments show that previous scaling rules fail in the training of CTR prediction neural networks. To tackle this problem, we first theoretically show that different frequencies of ids make it challenging to scale hyperparameters when scaling the batch size. To stabilize the training process in a large batch size setting, we develop the adaptive Column-wise Clipping (CowClip). It enables an easy and effective scaling rule for the embeddings, which keeps the learning rate unchanged and scales the L2 loss. We conduct extensive experiments with four CTR prediction networks on two real-world datasets and successfully scaled 128 times the original batch size without accuracy loss. In particular, for CTR prediction model DeepFM training on the Criteo dataset, our optimization framework enlarges the batch size from 1K to 128K with over 0.1% AUC improvement and reduces training time from 12 hours to 10 minutes on a single V100 GPU. Our code locates at github.com/bytedance/LargeBatchCTR.

Haochen Liu, Joseph Thekinen, Sinem Mollaoglu, Da Tang, Ji Yang, Youlong Cheng, Hui Liu, Jiliang Tang. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2022.

Real-world recommendation datasets have been shown to be subject to selection bias, which can challenge recommendation models to learn real preferences of users, so as to make accurate recommendations. Existing approaches to mitigate selection bias, such as data imputation and inverse propensity score, are sensitive to the quality of the additional imputation or propensity estimation models. To break these limitations, in this work, we propose a novel self-supervised learning (SSL) framework, i.e., Rating Distribution Calibration (RDC), to tackle selection bias without introducing additional models. In addition to the original training objective, we introduce a rating distribution calibration loss. It aims to correct the predicted rating distribution of biased users by taking advantage of that of their similar unbiased users. We empirically evaluate RDC on two real-world datasets and one synthetic dataset. The experimental results show that RDC outperforms the original model as well as the state-of-the-art debiasing approaches by a significant margin.

Variational Auto-Encoders (VAEs) are capable of learning latent representations for high dimensional data. However, due to the i.i.d. assumption, VAEs only optimize the singleton variational distributions and fail to account for the correlations between data points, which might be crucial for learning latent representations from dataset where a priori we know correlations exist. We propose Correlated Variational Auto-Encoders (CVAEs) that can take the correlation structure into consideration when learning latent representations with VAEs. CVAEs apply a prior based on the correlation structure. To address the intractability introduced by the correlated prior, we develop an approximation by average of a set of tractable lower bounds over all maximal acyclic subgraphs of the undirected correlation graph. Experimental results on matching and link prediction on public benchmark rating datasets and spectral clustering on a synthetic dataset show the effectiveness of the proposed method over baseline algorithms.

Variational inference transforms posterior inference into parametric optimization thereby enabling the use of latent variable models where otherwise impractical. However, variational inference can be finicky when different variational parameters control variables that are strongly correlated under the model. Traditional natural gradients based on the variational approximation fail to correct for correlations when the approximation is not the true posterior. To address this, we construct a new natural gradient called the Variational Predictive Natural Gradient (VPNG). Unlike traditional natural gradients for variational inference, this natural gradient accounts for the relationship between model parameters and variational parameters. We demonstrate the insight with a simple example as well as the empirical value on a classification task, a deep generative model of images, and probabilistic matrix factorization for recommendation.

Developing agents to engage in complex goaloriented dialogues is challenging partly because the main learning signals are very sparse in long conversations.In this paper, we propose a divide-and-conquer approach that discovers and exploits the hidden structure of the task to enable efficient policy learning.First, given successful example dialogues, we propose the Subgoal Discovery Network (SDN) to divide a complex goal-oriented task into a set of simpler subgoals in an unsupervised fashion.We then use these subgoals to learn a multi-level policy by hierarchical reinforcement learning.We demonstrate our method by building a dialogue agent for the composite task of travel planning.Experiments with simulated and real users show that our approach performs competitively against a state-of-theart method that requires human-defined subgoals.Moreover, we show that the learned subgoals are often human comprehensible.

We consider the problem of consistently matching multiple sets of elements to each other, which is a common task in fields such as computer vision. To solve the underlying NP-hard objective, existing methods often relax or approximate it, but end up with unsatisfying empirical performance due to a misaligned objective. We propose a coordinate update algorithm that directly optimizes the target objective. By using pairwise alignment information to build an undirected graph and initializing the permutation matrices along the edges of its Maximum Spanning Tree, our algorithm successfully avoids bad local optima. Theoretically, with high probability our algorithm guarantees an optimal solution under reasonable noise assumptions. Empirically, our algorithm consistently and significantly outperforms existing methods on several benchmark tasks on real datasets.

Mean-shift is an effective algorithm for object tracking. However, it has a poor performance when the illumination condition changes fast or the tracking target being shadowed. By contract, particle filter based object tracking has a better tracking performance, but the tracking speed is much slower compared to mean-shift. Owing to the limitations of just using a single algorithm, a novel object tracking method based on both mean-shift and particle filter is proposed in this paper. A system with feedback has been constructed by the proposed method, in which the mean-shift technique is used for initial registration, and the particle filter is called to improve the performance of tracking when the tracking result with mean-shift is unconvincing. RGB color histogram is exploited as image feature and Bhattacharyya coefficient is used for measuring the similarity between object model and candidate regions. Tracking experiments evaluated on various videos show that the proposed method is well-behaved in cases that objects have shift-variant, rotation and scaling, and achieves a satisfying tracking speed.