Search

Large foundation language models have shown their versatility in being able to be adapted to perform a wide variety of downstream tasks, such as text generation, sentiment analysis, semantic search etc. However, training such large foundational models is a non-trivial exercise that requires a significant amount of compute power and expertise from machine learning and systems experts. As models get larger, these demands are only increasing. Sparsity is a promising technique to relieve the compute requirements for training. However, sparsity introduces new challenges in training the sparse model to the same quality as the dense counterparts. Furthermore, sparsity drops the operation intensity and introduces irregular memory access patterns that makes it challenging to efficiently utilize compute resources. This paper demonstrates an end-to-end training flow on a large language model - 13 billion GPT - using sparsity and dataflow. The dataflow execution model and architecture enables efficient on-chip irregular memory accesses as well as native kernel fusion and pipelined parallelism that helps recover device utilization. We show that we can successfully train GPT 13B to the same quality as the dense GPT 13B model, while achieving an end-end speedup of 4.5x over dense A100 baseline.

Introducing nanorods into a polymeric matrix can enhance the physical and mechanical properties of the resulting material. In this paper, we focus on understanding the dispersion and orientation patterns of nanorods in an unentangled polymer melt, particularly as a function of nanorod concentration, using Molecular Dynamics (MD) simulations. The system is comprised of flexible polymer chains and multi-thread nanorods that are equilibrated in the NPT ensemble. All interactions are purely repulsive except for those between polymers and rods. Results with attractive versus repulsive polymer-rod interactions are compared and contrasted. The concentration of rods has a direct impact on the phase behaviour of the system. At lower concentrations rods phase separate into nematic clusters, while at higher concentrations more isotropic and less structured rod configurations are observed. A detailed examination of the conformation of the polymer chains near the rod surface shows extension of the chains along the director of the rods (especially within clusters). The dispersion and orientation of the nanorods is a result of the competition between depletion entropic forces responsible for the formation of rod clusters, the enthalpic effects that improve mixing of rods and polymer, and entropic losses of polymers interpenetrating rod clusters., Comment: 16 pages, 15 figures

Physically relevant field-theoretic quantities are usually derived from perturbation techniques. These quantities are solved in the form of an asymptotic series in powers of small perturbation parameters related to the physical system, and calculating higher powers typically results in a higher amount of computational complexity. Such divergent expansions were handled using hyper-geometric functions, continued functions, and Borel-Leroy transforms. Hypergeometric functions are expanded as series, and a rough estimate of next-order information is predicted using information from known orders. Continued functions are used for the resummation of these series. The effective nature of extrapolation using such functions is illustrated by taking different examples in field theories. In the vicinity of second-order phase transitions, critical exponents are the most interesting numerical quantities corresponding to a wide range of physical systems. Using the techniques mentioned in this work, precise estimates are obtained for these critical exponents in {\phi}^4 and {\phi}^3 field models.

Scientific workflows have become integral tools in broad scientific computing use cases. Science discovery is increasingly dependent on workflows to orchestrate large and complex scientific experiments that range from execution of a cloud-based data preprocessing pipeline to multi-facility instrument-to-edge-to-HPC computational workflows. Given the changing landscape of scientific computing and the evolving needs of emerging scientific applications, it is paramount that the development of novel scientific workflows and system functionalities seek to increase the efficiency, resilience, and pervasiveness of existing systems and applications. Specifically, the proliferation of machine learning/artificial intelligence (ML/AI) workflows, need for processing large scale datasets produced by instruments at the edge, intensification of near real-time data processing, support for long-term experiment campaigns, and emergence of quantum computing as an adjunct to HPC, have significantly changed the functional and operational requirements of workflow systems. Workflow systems now need to, for example, support data streams from the edge-to-cloud-to-HPC enable the management of many small-sized files, allow data reduction while ensuring high accuracy, orchestrate distributed services (workflows, instruments, data movement, provenance, publication, etc.) across computing and user facilities, among others. Further, to accelerate science, it is also necessary that these systems implement specifications/standards and APIs for seamless (horizontal and vertical) integration between systems and applications, as well as enabling the publication of workflows and their associated products according to the FAIR principles. This document reports on discussions and findings from the 2022 international edition of the Workflows Community Summit that took place on November 29 and 30, 2022.

Reverse engineering of FPGA based designs from the flattened LUT level netlist to high level RTL helps in verification of the design or in understanding legacy designs. We focus on flattened netlists for FPGA devices from Xilinx 7 series and Zynq 7000. We propose a design element grouping algorithm that makes use of the location information of the elements on the physical device after place and route. The proposed grouping algorithm gives clusters with average NMI of 0.73 for groupings including all element types. The benchmarks chosen include a range of designs from communication, arithmetic units, processors and DSP processing units., Comment: Paper accepted into proceedings of VLSID2023 conference

Reverse engineering of FPGA designs from bitstreams to RTL models aids in understanding the high level functionality of the design and for validating and reconstructing legacy designs. Fast carry-chains are commonly used in synthesis of operators in FPGA designs. We propose a method to detect word-level structures by analyzing these carry-chains in LUT (Look-Up Table) level netlists. We also present methods to adapt existing techniques to identify combinational operations and sequential modules in ASIC netlists to LUT netlists. All developed and adapted techniques are consolidated into an integrated tool-chain to aid in reverse engineering of word-level designs from LUT-level netlists. When evaluated on a set of real-world designs, the tool-chain infers 34\% to 100\% of the elements in the netlist to be part of a known word-level operation or a known sequential module., Comment: 8 pages, 6 figures, Accepted to appear in ISQED 2023 conference

This research aims to examine the usefulness of integrating various feature selection methods with regression algorithms for sleep quality prediction. A publicly accessible sleep quality dataset is used to analyze the effect of different feature selection techniques on the performance of four regression algorithms - Linear regression, Ridge regression, Lasso Regression and Random Forest Regressor. The results are compared to determine the optimal combination of feature selection techniques and regression algorithms. The conclusion of the study enriches the current literature on using machine learning for sleep quality prediction and has practical significance for personalizing sleep recommendations for individuals.

Quantum decoherence happens when the system interacts with the environment. Quantum correlation behaviours in the two-qubit spin squeezing model are studied under the influence of intrinsic decoherence. Quantitative results were determined, which depend on parameters of the physical system by checking different quantifiers of quantum correlation such as entanglement, local quantum uncertainty, trace distance discord and uncertainty-induced quantum nonlocality. We show that the entanglement suffers from intrinsic decoherence and exhibits sudden death, whereas the other measures are more robust against intrinsic decoherence. Further, we highlight the role of spin squeezing coupling constant and magnetic field., Comment: 10 pages

Resummation methods using continued functions are implemented to converge divergent series appearing in perturbation problems related to continuous phase transitions in field theories. In some cases, better convergence properties are obtained using continued functions than diagonal Pade approximants, which are extensively used in literature. We check the reliability of critical exponent estimates derived previously in universality classes of O(n)-symmetric models (classical phase transitions) and Gross-Neveu-Yukawa models (quantum phase transitions) using new methods., Comment: 22 pages

This paper deals with the development of a localization methodology for autonomous vehicles using only a $3\Dim$ LIDAR sensor. In the context of this paper, localizing a vehicle in a known 3D global map of the environment is essentially to find its global $3\Dim$ pose (position and orientation) within this map. The problem of tracking is then to use sequential LIDAR scan measurement to also estimate other states such as velocity and angular rates, in addition to the global pose of the vehicle. Particle filters are often used in localization and tracking, as in applications of simultaneously localization and mapping. But particle filters become computationally prohibitive with the increase in particles, often required to localize in a large $3\Dim$ map. Further, computing the likelihood of a LIDAR scan for each particle is in itself a computationally expensive task, thus limiting the number of particles that can be used for real time performance. To this end, we propose a hybrid approach that combines the advantages of a particle filter with a global-local scan matching method to better inform the re-sampling stage of the particle filter. Further, we propose to use a pre-computed likelihood grid to speedup the computation of LIDAR scans. Finally, we develop the complete algorithm to extensively leverage parallel processing to achieve near sufficient real-time performance on publicly available KITTI datasets.

Modern-day autonomous vehicles are increasingly becoming complex multidisciplinary systems composed of mechanical, electrical, electronic, computing and information sub-systems. Furthermore, the individual constituent technologies employed for developing autonomous vehicles have started maturing up to a point, where it seems beneficial to start looking at the synergistic integration of these components into sub-systems, systems, and potentially, system-of-systems. Hence, this work applies the principles of mechatronics approach of system design, verification and validation for the development of autonomous vehicles. Particularly, we discuss leveraging multidisciplinary co-design practices along with virtual, hybrid and physical prototyping and testing within a concurrent engineering framework to develop and validate a scaled autonomous vehicle using the AutoDRIVE ecosystem. We also describe a case-study of autonomous parking application using a modular probabilistic framework to illustrate the benefits of the proposed approach., Comment: Accepted at IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM) 2023

Many modern applications involve accessing and processing graphical data, i.e. data that is naturally indexed by graphs. Examples come from internet graphs, social networks, genomics and proteomics, and other sources. The typically large size of such data motivates seeking efficient ways for its compression and decompression. The current compression methods are usually tailored to specific models, or do not provide theoretical guarantees. In this paper, we introduce a low-complexity lossless compression algorithm for sparse marked graphs, i.e. graphical data indexed by sparse graphs, which is capable of universally achieving the optimal compression rate in a precisely defined sense. In order to define universality, we employ the framework of local weak convergence, which allows one to make sense of a notion of stochastic processes for sparse graphs. Moreover, we investigate the performance of our algorithm through some experimental results on both synthetic and real-world data.

We tackle the task of synthesizing novel views of an object given a few input images and associated camera viewpoints. Our work is inspired by recent 'geometry-free' approaches where multi-view images are encoded as a (global) set-latent representation, which is then used to predict the color for arbitrary query rays. While this representation yields (coarsely) accurate images corresponding to novel viewpoints, the lack of geometric reasoning limits the quality of these outputs. To overcome this limitation, we propose 'Geometry-biased Transformers' (GBTs) that incorporate geometric inductive biases in the set-latent representation-based inference to encourage multi-view geometric consistency. We induce the geometric bias by augmenting the dot-product attention mechanism to also incorporate 3D distances between rays associated with tokens as a learnable bias. We find that this, along with camera-aware embeddings as input, allows our models to generate significantly more accurate outputs. We validate our approach on the real-world CO3D dataset, where we train our system over 10 categories and evaluate its view-synthesis ability for novel objects as well as unseen categories. We empirically validate the benefits of the proposed geometric biases and show that our approach significantly improves over prior works., Comment: Project page: https://mayankgrwl97.github.io/gbt

Performance verification is a nascent but promising tool for understanding the performance and limitations of heuristics under realistic assumptions. Bespoke performance verification tools have already demonstrated their value in settings like congestion control and packet scheduling. In this paper, we aim to emphasize the broad applicability and utility of performance verification. To that end, we highlight the design principles of performance verification. Then, we leverage that understanding to develop a set of easy-to-follow guidelines that are applicable to a wide range of resource allocation heuristics. In particular, we introduce Virelay, a framework that enables heuristic designers to express the behavior of their algorithms and their assumptions about the system in an environment that resembles a discrete-event simulator. We demonstrate the utility and ease-of-use of Virelay by applying it to six diverse case studies. We produce bounds on the performance of classical algorithms, work stealing and SRPT scheduling, under practical assumptions. We demonstrate Virelay's expressiveness by capturing existing models for congestion control and packet scheduling, and we verify the observation that TCP unfairness can cause some ML training workloads to spontaneously converge to a state of high network utilization. Finally, we use Virelay to identify two bugs in the Linux CFS load balancer., Comment: 12 pages, 11 figures

In optimization-based approaches to inverse problems and to statistical estimation, it is common to augment criteria that enforce data fidelity with a regularizer that promotes desired structural properties in the solution. The choice of a suitable regularizer is typically driven by a combination of prior domain information and computational considerations. Convex regularizers are attractive computationally but they are limited in the types of structure they can promote. On the other hand, nonconvex regularizers are more flexible in the forms of structure they can promote and they have showcased strong empirical performance in some applications, but they come with the computational challenge of solving the associated optimization problems. In this paper, we seek a systematic understanding of the power and the limitations of convex regularization by investigating the following questions: Given a distribution, what is the optimal regularizer for data drawn from the distribution? What properties of a data source govern whether the optimal regularizer is convex? We address these questions for the class of regularizers specified by functionals that are continuous, positively homogeneous, and positive away from the origin. We say that a regularizer is optimal for a data distribution if the Gibbs density with energy given by the regularizer maximizes the population likelihood (or equivalently, minimizes cross-entropy loss) over all regularizer-induced Gibbs densities. As the regularizers we consider are in one-to-one correspondence with star bodies, we leverage dual Brunn-Minkowski theory to show that a radial function derived from a data distribution is akin to a ``computational sufficient statistic'' as it is the key quantity for identifying optimal regularizers and for assessing the amenability of a data source to convex regularization.

The lithium metal-solid-state electrolyte interface plays a critical role in the performance of solid-state batteries. However, operando characterization of the buried interface morphology in solid-state cells is particularly difficult because of the lack of direct optical access. Destructive techniques that require isolating the interface inadvertently modify the interface and cannot be used for operando monitoring. In this work, we introduce the concept of thermal wave sensing using modified 3-omega sensors that is attached to the outside of the lithium metal-solid state cells to non-invasively probe the morphology of the lithium metal-electrolyte interface. We show that the thermal interface resistance measured by the 3-omega sensors relates directly to the physical morphology of the interface and demonstrate that 3-omega thermal wave sensing can be used for non-invasive operando monitoring the morphology evolution of the lithium metal-solid state electrolyte interface.

In this work, we give efficient algorithms for privately estimating a Gaussian distribution in both pure and approximate differential privacy (DP) models with optimal dependence on the dimension in the sample complexity. In the pure DP setting, we give an efficient algorithm that estimates an unknown $d$-dimensional Gaussian distribution up to an arbitrary tiny total variation error using $\widetilde{O}(d^2 \log \kappa)$ samples while tolerating a constant fraction of adversarial outliers. Here, $\kappa$ is the condition number of the target covariance matrix. The sample bound matches best non-private estimators in the dependence on the dimension (up to a polylogarithmic factor). We prove a new lower bound on differentially private covariance estimation to show that the dependence on the condition number $\kappa$ in the above sample bound is also tight. Prior to our work, only identifiability results (yielding inefficient super-polynomial time algorithms) were known for the problem. In the approximate DP setting, we give an efficient algorithm to estimate an unknown Gaussian distribution up to an arbitrarily tiny total variation error using $\widetilde{O}(d^2)$ samples while tolerating a constant fraction of adversarial outliers. Prior to our work, all efficient approximate DP algorithms incurred a super-quadratic sample cost or were not outlier-robust. For the special case of mean estimation, our algorithm achieves the optimal sample complexity of $\widetilde O(d)$, improving on a $\widetilde O(d^{1.5})$ bound from prior work. Our pure DP algorithm relies on a recursive private preconditioning subroutine that utilizes the recent work on private mean estimation [Hopkins et al., 2022]. Our approximate DP algorithms are based on a substantial upgrade of the method of stabilizing convex relaxations introduced in [Kothari et al., 2022].

Prototyping and validating hardware-software components, sub-systems and systems within the intelligent transportation system-of-systems framework requires a modular yet flexible and open-access ecosystem. This work presents our attempt towards developing such a comprehensive research and education ecosystem, called AutoDRIVE, for synergistically prototyping, simulating and deploying cyber-physical solutions pertaining to autonomous driving as well as smart city management. AutoDRIVE features both software as well as hardware-in-the-loop testing interfaces with openly accessible scaled vehicle and infrastructure components. The ecosystem is compatible with a variety of development frameworks, and supports both single and multi-agent paradigms through local as well as distributed computing. Most critically, AutoDRIVE is intended to be modularly expandable to explore emergent technologies, and this work highlights various complementary features and capabilities of the proposed ecosystem by demonstrating four such deployment use-cases: (i) autonomous parking using probabilistic robotics approach for mapping, localization, path planning and control; (ii) behavioral cloning using computer vision and deep imitation learning; (iii) intersection traversal using vehicle-to-vehicle communication and deep reinforcement learning; and (iv) smart city management using vehicle-to-infrastructure communication and internet-of-things.

Neural networks have revolutionized the area of artificial intelligence and introduced transformative applications to almost every scientific field and industry. However, this success comes at a great price; the energy requirements for training advanced models are unsustainable. One promising way to address this pressing issue is by developing low-energy neuromorphic hardware that directly supports the algorithm's requirements. The intrinsic non-volatility, non-linearity, and memory of spintronic devices make them appealing candidates for neuromorphic devices. Here we focus on the reservoir computing paradigm, a recurrent network with a simple training algorithm suitable for computation with spintronic devices since they can provide the properties of non-linearity and memory. We review technologies and methods for developing neuromorphic spintronic devices and conclude with critical open issues to address before such devices become widely used.

In this paper, some adaptive single-step methods like Trapezoid (TR), Implicit-mid point (IMP), Euler-backward (EB), and Radau IIA (Rad) methods are implemented in Maple to solve index-1 nonlinear Differential Algebraic Equations (DAEs). Maple's robust and efficient ability to search within a list/set is exploited to identify the sparsity pattern and the analytic Jacobian. The algorithm and implementation were found to be robust and efficient for index-1 DAE problems and scales well for finite difference/finite element discretization of two-dimensional models with system size up to 10,000 nonlinear DAEs and solves the same in few seconds.