Speeds up deep neural networks training by 10x using GPU concurrency.
problem Training deep residual neural networks efficiently.
method Layer-wise parallel training with GPU concurrency and Nonlinear Multigrid.
result 10.2x speedup over traditional techniques.
GPUPoly verifies large neural networks robustly on GPUs.
problem Proving robustness of deep neural networks is crucial but challenging.
method Custom polyhedra algorithms on GPUs.
result GPUPoly can verify 1M neuron networks in 34.5 ms.
The paper optimizes neural network inference on mobile GPUs.
problem Limited computing power and thermal constraints on mobile CPUs.
method Leverage mobile GPUs for neural network inference.
result Real-time inference of deep neural networks on Android and iOS devices.
MoGA optimizes mobile networks for GPU efficiency.
problem Lack of GPU-aware optimization in mobile network design.
method Proposes MoGA, a GPU-aware neural architecture search.
result Achieves higher accuracy with similar latency constraints.
Study shows Transformer and Neural GPU are Turing complete without external memory.
problem Exploring computational power of modern neural network architectures.
method Analyzing computational properties of Transformer and Neural GPU.
result Transformer and Neural GPU are Turing complete without external memory.
Paper proposes a GPU-based system for training massive deep learning models in ads systems.
problem Training massive deep learning models with terabyte-scale parameters in ads systems.
method Hierarchical GPU parameter server with 3-layer storage (GPU High-Bandwidth Memory, CPU main memory, SSD).
result 4-node hierarchical GPU parameter server trains a model 2X faster than a 150-node in-memory system.
New GPU kernels boost deep learning speed and memory efficiency.
problem Sparse deep learning matrices are not well-suited for existing sparse kernels.
method Identified favorable properties of sparse matrices from deep learning, developed high-performance GPU kernels for sparse matrix operations.
result 27% of single-precision peak performance on Nvidia V100 GPUs achieved with new kernels.
One reflection suffices for orthogonal weights, reducing GPU usage.
problem Efficiently computing orthogonal weight matrices without high GPU utilization.
method Use an auxiliary neural network to compute one reflection instead of many.
result One reflection is sufficient for orthogonal weights, improving GPU utilization.
GPU-accelerated particle methods outperform neural samplers in LFT benchmarks.
problem High-dimensional multimodal sampling problems in lattice field theory.
method GPU-accelerated particle Monte Carlo methods (Sequential Monte Carlo and nested sampling).
result These methods match or outperform neural samplers in sample quality and wall-clock time.
Differentiable voxelization for 3D meshes with GPU acceleration.
problem Efficient and accurate voxelization of 3D meshes.
method Differentiable voxelization using winding number and solid angles, with GPU acceleration and neural network deformation.
result State-of-the-art performance in accuracy and efficiency on the ShapeNet dataset.
The Long-Short-Term-Memory Recurrent Neural Networks (LSTM RNNs) are a popular class of machine learning models for analyzing sequential data. Their training on modern GPUs, however, is limited by the GPU memory capacity. Our profiling results of the LSTM RNN-based Neural Machine Translation (NMT) model reveal that fea…
Neural architecture search (NAS) has a great impact by automatically designing effective neural network architectures. However, the prohibitive computational demand of conventional NAS algorithms (e.g. 104 GPU hours) makes it difficult to \emph{directly} search the architectures on large-scale tasks (e.g. ImageNet).…
New algorithms reduce communication in GNN training.
problem Higher communication costs in GNNs due to sparse connectivity.
method Parallel algorithms for sparse-dense matrix multiplication.
result Asymptotic reduction in communication compared to previous methods.
This paper accelerates sparse CNN layers on GPUs by using unstructured sparsity.
problem Efficiency of sparse CNN layers on GPUs.
method Direct sparse operation and reduced precision.
result Achieving up to 90% sparsity in deep CNN models improves efficiency.
SSH-Net: A Deep Neural Network for Predicting Failure Time Distribution Functions under Competing Risks with GPU Data
problem Predicting failure time distribution functions under competing risks
method Structured Segmented Hazard Deep Neural Network (SSH-Net)
result Prediction accuracy validated through simulation studies and GPU data
Benchmarking TPU, GPU, and CPU for deep learning models.
problem Improving performance in deep learning training.
method ParaDnn benchmark suite for FC, CNN, and RNN models on TPU, GPU, and CPU.
result TPU, GPU, and CPU have unique strengths for different types of models.
Performance-aware channel pruning improves CNN on embedded GPUs.
problem Inefficient channel pruning on embedded GPUs leads to performance slowdowns.
method Evaluate higher-level libraries that analyze input characteristics for optimized code generation.
result Performance-aware pruning can achieve significant performance speedups, up to 10x.
HotNAS reduces AI search time from hundreds of GPU hours to less than 3 GPU hours.
problem High time required for AI solution generation from scratch.
method HotNAS starts from a 'hot' state using pre-trained models and integrates compression during co-search.
result Reduces search time from 200 GPU hours to less than 3 GPU hours.
This paper optimizes neural network training by packing multiple models on a single GPU.
problem Efficiently sharing limited training resources among multiple neural network models.
method Proposes a primitive called 'pack' to jointly train multiple models on a single GPU.
result Significant performance improvements for hyperparameter tuning, up to 40% for two models.
Neural machine translation used to convert CUDA to OpenCL.
problem Translating CUDA to OpenCL programs.
method Training input set generation, pre/post processing, case study.
result Improved accuracy in translating CUDA to OpenCL.
Convolutional neural networks have recently achieved significant breakthroughs in various image classification tasks. However, they are computationally expensive,which can make their feasible mplementation on embedded and low-power devices difficult. In this paper convolutional neural network binarization is implemente…
SparseRT accelerates sparse computations on GPUs for deep learning inference.
problem Efficiently handling unstructured sparsity patterns on GPUs for deep learning.
method SparseRT, a code generator that leverages unstructured sparsity for accelerating sparse linear algebra operations.
result Geometric mean speedups of 3.4x at 90% sparsity and 5.4x at 95% sparsity for 1x1 convolutions and fully connected layers.
DistShap parallelizes GNN explanation for large graphs.
problem Computational expense in attributing GNN predictions to specific edges or features.
method Distributed Shapley values across multiple GPUs for scalable GNN explanations.
result DistShap outperforms existing methods and scales to models with millions of features.
Investigates the impact of batch size on GPU and TPU performance.
problem Optimizing performance of GPUs and TPUs during training and inference phases.
method Investigated the impact of batch size on performance of GPUs and TPUs using standard MNIST and Fashion-MNIST datasets.
result Significant speedup was achieved even with low-scale usage of TPUv2 units, up to 10x for training and 2x for prediction.
FPGAs accelerate DNNs with binarized weights, improving power and speed.
problem Power and speed limitations of GPU-accelerated DNNs.
method Binarized neural networks on FPGAs using OpenCL.
result Near state-of-the-art performance with >16x power savings.
A new method increases 3D medical image segmentation accuracy and speed.
problem Training large 3D medical images on GPUs with limited memory.
method Data-swapping method to enlarge GPU memory and avoid patching.
result Improved segmentation accuracy and speed for full-size images.
Neural architecture search (NAS) has been proposed to automatically tune deep neural networks, but existing search algorithms, e.g., NASNet, PNAS, usually suffer from expensive computational cost. Network morphism, which keeps the functionality of a neural network while changing its neural architecture, could be helpfu…
A GPU-based workflow for building physics emulators of hypersonic flows
problem Resolving complex physical phenomena in hypersonic flows
method Fully GPU-based workflow integrating accelerated data generation and neural emulators
result Physics emulators remain reliable beyond their training distribution
MIOpen optimizes deep learning operators for GPUs, accelerating research.
problem Optimizing deep learning operators for efficient GPU execution.
method Highly optimized implementations, fusion, auto-tuning, bfloat16 support.
result Accelerates time to discovery in deep learning research.
Proposes RBGP framework for efficient block sparse neural networks.
problem Efficiently exploit structured sparsity patterns for sparse neural networks on GPU.
method Uses Ramanujan Bipartite Graph Product to generate structured multi-level block sparse neural networks.
result Achieves 5-9x and 2-5x runtime gains over unstructured and block sparsity patterns respectively, while maintaining accuracy.
New kernels boost RNN performance on non-time-series data.
problem Improving performance of RNNs on non-time-series data.
method Extended RNN kernels to complex architectures, developed fast GPU implementation.
result RNN-based classifiers outperform baselines on 90 non-time-series datasets.
Synchronized stochastic gradient descent (SGD) optimizers with data parallelism are widely used in training large-scale deep neural networks. Although using larger mini-batch sizes can improve the system scalability by reducing the communication-to-computation ratio, it may hurt the generalization ability of the models…
New GPU algorithm boosts machine learning with larger datasets.
problem Limited GPU memory restricts training data size.
method Out-of-core GPU gradient boosting algorithm.
result Training larger datasets on GPUs without accuracy loss.
The design of neural network architectures for a new data set is a laborious task which requires human deep learning expertise. In order to make deep learning available for a broader audience, automated methods for finding a neural network architecture are vital. Recently proposed methods can already achieve human expe…
Massively parallel architectures such as the GPU are becoming increasingly important due to the recent proliferation of data. In this paper, we propose a key class of hybrid parallel graphlet algorithms that leverages multiple CPUs and GPUs simultaneously for computing k-vertex induced subgraph statistics (called graph…
End-to-end speech recognition system trained on GPUs and CPUs.
problem Building state-of-the-art speech recognition systems.
method Utilizes CPUs and GPUs for training, data augmentation, and neural network updates. Uses vocal tract length perturbation and acoustic simulator for data augmentation. Employed Horovod allreduce for training.
result Achieved 7.92% WER on proprietary English Bixby open domain test set using a Bidirectional Full Attention (BFA) model.
CuLE accelerates reinforcement learning on GPUs.
problem Limited CPU-GPU communication bandwidth in Atari Learning Environment.
method CUDA port of ALE, leveraging GPU parallelization and direct frame rendering.
result Up to 155M frames per hour on a single GPU, accelerating reinforcement learning.
Extends GENO framework for GPU optimization of constrained ML problems.
problem Constrained optimization in classical machine learning.
method Extends GENO framework to GPU optimization, specifying problems in a modeling language.
result Solvers on GPU outperform state-of-the-art approaches by several orders of magnitude.
GPU computing has become popular in computational finance and many financial institutions are moving their CPU based applications to the GPU platform. Since most Monte Carlo algorithms are embarrassingly parallel, they benefit greatly from parallel implementations, and consequently Monte Carlo has become a focal point …
Deep learning improves gamma-ray energy estimation and event selection.
problem Improving gamma-ray event selection and energy estimation.
method Adapted convolutional neural networks (CNN) for gamma-ray astronomy.
result Significant improvement in gamma-ray energy estimation and event selection.
GPU speeds up derivatives sensitivity computation for Heston model options.
problem Efficient computation of option Greeks under the Heston model.
method Implemented exact simulation and novel Milstein discretisation methods on GPU.
result GPU speeds up Greeks computation up to 200x compared to CPU methods.
Study feasibility of deep neural networks for Euro banknote classification.
problem Meeting special requirements for banknote classification by central banks.
method Training and testing deep neural networks on state-of-the-art GPU hardware for few classes and 0-class rejection.
result Deep neural networks can meet central bank requirements for banknote classification.
Convolutional Neural Networks (CNN) has become more popular choice for various tasks such as computer vision, speech recognition and natural language processing. Thanks to their large computational capability and throughput, GPUs ,which are not power efficient and therefore does not suit low power systems such as mobil…
JAXFit speeds up curve fitting on GPUs.
problem Nonlinear least squares curve fitting problems.
method Trust region method on GPU with automatic differentiation.
result Significantly faster than CPU and other GPU libraries.
New theory maps neural network weights to optimize faster and scale.
problem Optimizing neural networks for speed and scalability.
method Constructing a duality map using layer-wise operator norms.
result Derived GPU-friendly algorithms for various layers.
GPU optimization speeds up large-scale classification tasks.
problem Efficiently training large-scale classification models on GPUs.
method Judecious GPU-optimization principles applied to TRON algorithm.
result Significant speedups for logistic regression and SVM classification.
A GPU framework speeds up BnB for discrete optimization problems.
problem Optimizing large-scale discrete problems with GPU limitations.
method Parallel BnB nodes in GPU batches, using padding and custom kernels.
result One to two orders of magnitude speedup and zero optimality gap.
Efficiently optimizes orthogonal and Stiefel matrices on parallel units.
problem Optimization over orthogonal groups on parallel units.
method CWY and T-CWY transforms for parametrization and optimization.
result CWY and T-CWY methods lead to convergence on parallel units.