FPGA technology mapping is the process of implementing a hardware design expressed in high-level HDL (hardware design language) code using the low-level, architecture-specific primitives of the target FPGA. As FPGAs become increasingly heterogeneous, achieving high performance requires hardware synthesis tools that better support mapping to complex, highly configurable primitives like digital signal processors (DSPs). Current tools support DSP mapping via handwritten special-case mapping rules, which are laborious to write, error-prone, and often overlook mapping opportunities. We introduce Lakeroad, a principled approach to technology mapping via sketch-guided program synthesis. Lakeroad leverages two techniques -- architecture-independent sketch templates and semantics extraction from HDL -- to provide extensible technology mapping with stronger correctness guarantees and higher coverage of mapping opportunities than state-of-the-art tools. Across representative microbenchmarks, Lakeroad produces 2--3.5$\times$ the number of optimal mappings compared to proprietary state-of-the-art tools and 6--44$\times$ the number of optimal mappings compared to popular open-source tools, while also providing correctness guarantees not given by any other tool.
相關內容
Large-scale datasets are important for the development of deep learning models. Such datasets usually require a heavy workload of annotations, which are extremely time-consuming and expensive. To accelerate the annotation procedure, multiple annotators may be employed to label different subsets of the data. However, the inconsistency and bias among different annotators are harmful to the model training, especially for qualitative and subjective tasks.To address this challenge, in this paper, we propose a novel contrastive regression framework to address the disjoint annotations problem, where each sample is labeled by only one annotator and multiple annotators work on disjoint subsets of the data. To take account of both the intra-annotator consistency and inter-annotator inconsistency, two strategies are employed.Firstly, a contrastive-based loss is applied to learn the relative ranking among different samples of the same annotator, with the assumption that the ranking of samples from the same annotator is unanimous. Secondly, we apply the gradient reversal layer to learn robust representations that are invariant to different annotators. Experiments on the facial expression prediction task, as well as the image quality assessment task, verify the effectiveness of our proposed framework.
Neural architecture search (NAS) has become a key component of AutoML and a standard tool to automate the design of deep neural networks. Recently, training-free NAS as an emerging paradigm has successfully reduced the search costs of standard training-based NAS by estimating the true architecture performance with only training-free metrics. Nevertheless, the estimation ability of these metrics typically varies across different tasks, making it challenging to achieve robust and consistently good search performance on diverse tasks with only a single training-free metric. Meanwhile, the estimation gap between training-free metrics and the true architecture performances limits training-free NAS to achieve superior performance. To address these challenges, we propose the robustifying and boosting training-free NAS (RoBoT) algorithm which (a) employs the optimized combination of existing training-free metrics explored from Bayesian optimization to develop a robust and consistently better-performing metric on diverse tasks, and (b) applies greedy search, i.e., the exploitation, on the newly developed metric to bridge the aforementioned gap and consequently to boost the search performance of standard training-free NAS further. Remarkably, the expected performance of our RoBoT can be theoretically guaranteed, which improves over the existing training-free NAS under mild conditions with additional interesting insights. Our extensive experiments on various NAS benchmark tasks yield substantial empirical evidence to support our theoretical results.
In the realm of Business Process Management (BPM), process modeling plays a crucial role in translating complex process dynamics into comprehensible visual representations, facilitating the understanding, analysis, improvement, and automation of organizational processes. Traditional process modeling methods often require extensive expertise and can be time-consuming. This paper explores the integration of Large Language Models (LLMs) into process modeling to enhance flexibility, efficiency, and accessibility of process modeling for both expert and non-expert users. We propose a framework that leverages LLMs for the automated generation and iterative refinement of process models starting from textual descriptions. Our framework involves innovative prompting strategies for effective LLM utilization, along with a secure model generation protocol and an error-handling mechanism. Moreover, we instantiate a concrete system extending our framework. This system provides robust quality guarantees on the models generated and supports exporting them in standard modeling notations, such as the Business Process Modeling Notation (BPMN) and Petri nets. Preliminary results demonstrate the framework's ability to streamline process modeling tasks, underscoring the transformative potential of generative AI in the BPM field.
Scheduling distributed applications modeled as directed, acyclic task graphs to run on heterogeneous compute networks is a fundamental (NP-Hard) problem in distributed computing for which many heuristic algorithms have been proposed over the past decades. Many of these algorithms fall under the list-scheduling paradigm, whereby the algorithm first computes priorities for the tasks and then schedules them greedily to the compute node that minimizes some cost function. Thus, many algorithms differ from each other only in a few key components (e.g., the way they prioritize tasks, their cost functions, where the algorithms consider inserting tasks into a partially complete schedule, etc.). In this paper, we propose a generalized parametric list-scheduling algorithm that allows mixing and matching different algorithmic components to produce 72 unique algorithms. We benchmark these algorithms on four datasets to study the individual and combined effects of different algorithmic components on performance and runtime.
With the rapid development of deep learning models and hardware support for dense computing, the deep learning workload characteristics changed significantly from a few hot spots on compute-intensive operations to a broad range of operations scattered across the models. Accelerating a few compute-intensive operations using the expert-tuned implementation of primitives does not fully exploit the performance potential of AI hardware. Various efforts have been made to compile a full deep neural network (DNN) graph. One of the biggest challenges is to achieve high-performance tensor compilation by generating expert level performance code for the dense compute-intensive operations and applying compilation optimization at the scope of DNN computation graph across multiple compute-intensive operations. We present oneDNN Graph Compiler, a tensor compiler that employs a hybrid approach of using techniques from both compiler optimization and expert-tuned kernels for high performance code generation of the deep neural network graph. oneDNN Graph Compiler addresses unique optimization challenges in the deep learning domain, such as low-precision computation, aggressive fusion of graph operations, optimization for static tensor shapes and memory layout, constant weight optimization, and memory buffer reuse. Experimental results demonstrate significant performance gains over existing tensor compiler and primitives library for performance-critical DNN computation graphs and end-to-end models on Intel Xeon Scalable Processors.
With the advent of exascale computing, effective load balancing in massively parallel software applications is critically important for leveraging the full potential of high performance computing systems. Load balancing is the distribution of computational work between available processors. Here, we investigate the application of quantum annealing to load balance two paradigmatic algorithms in high performance computing. Namely, adaptive mesh refinement and smoothed particle hydrodynamics are chosen as representative grid and off-grid target applications. While the methodology for obtaining real simulation data to partition is application specific, the proposed balancing protocol itself remains completely general. In a grid based context, quantum annealing is found to outperform classical methods such as the round robin protocol but lacks a decisive advantage over more advanced methods such as steepest descent or simulated annealing despite remaining competitive. The primary obstacle to scalability is found to be limited coupling on current quantum annealing hardware. However, for the more complex particle formulation, approached as a multi-objective optimization, quantum annealing solutions are demonstrably Pareto dominant to state of the art classical methods across both objectives. This signals a noteworthy advancement in solution quality which can have a large impact on effective CPU usage.
While the study of language as typed on smartphones offers valuable insights, existing data collection methods often fall short in providing contextual information and ensuring user privacy. We present a privacy-respectful approach - context-enriched keyboard logging - that allows for the extraction of contextual information on the user's input motive, which is meaningful for linguistics, psychology, and behavioral sciences. In particular, with our approach, we enable distinguishing language contents by their channel (i.e., comments, messaging, search inputs). Filtering by channel allows for better pre-selection of data, which is in the interest of researchers and improves users' privacy. We demonstrate our approach on a large-scale six-month user study (N=624) of language use in smartphone interactions in the wild. Finally, we highlight the implications for research on language use in human-computer interaction and interdisciplinary contexts.
We propose a hybrid model predictive control algorithm, consensus complementarity control (C3), for systems that make and break contact with their environment. Many state-of-the-art controllers for tasks which require initiating contact with the environment, such as locomotion and manipulation, require a priori mode schedules or are too computationally complex to run at real-time rates. We present a method based on the alternating direction method of multipliers (ADMM) that is capable of high-speed reasoning over potential contact events. Via a consensus formulation, our approach enables parallelization of the contact scheduling problem. We validate our results on five numerical examples, including four high-dimensional frictional contact problems, and a physical experimentation on an underactuated multi-contact system. We further demonstrate the effectiveness of our method on a physical experiment accomplishing a high-dimensional, multi-contact manipulation task with a robot arm.
Learning latent representations of nodes in graphs is an important and ubiquitous task with widespread applications such as link prediction, node classification, and graph visualization. Previous methods on graph representation learning mainly focus on static graphs, however, many real-world graphs are dynamic and evolve over time. In this paper, we present Dynamic Self-Attention Network (DySAT), a novel neural architecture that operates on dynamic graphs and learns node representations that capture both structural properties and temporal evolutionary patterns. Specifically, DySAT computes node representations by jointly employing self-attention layers along two dimensions: structural neighborhood and temporal dynamics. We conduct link prediction experiments on two classes of graphs: communication networks and bipartite rating networks. Our experimental results show that DySAT has a significant performance gain over several different state-of-the-art graph embedding baselines.
Providing model-generated explanations in recommender systems is important to user experience. State-of-the-art recommendation algorithms -- especially the collaborative filtering (CF) based approaches with shallow or deep models -- usually work with various unstructured information sources for recommendation, such as textual reviews, visual images, and various implicit or explicit feedbacks. Though structured knowledge bases were considered in content-based approaches, they have been largely ignored recently due to the availability of vast amount of data and the learning power of many complex models. However, structured knowledge bases exhibit unique advantages in personalized recommendation systems. When the explicit knowledge about users and items is considered for recommendation, the system could provide highly customized recommendations based on users' historical behaviors and the knowledge is helpful for providing informed explanations regarding the recommended items. In this work, we propose to reason over knowledge base embeddings for explainable recommendation. Specifically, we propose a knowledge base representation learning framework to embed heterogeneous entities for recommendation, and based on the embedded knowledge base, a soft matching algorithm is proposed to generate personalized explanations for the recommended items. Experimental results on real-world e-commerce datasets verified the superior recommendation performance and the explainability power of our approach compared with state-of-the-art baselines.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191