Search Results for author:
Found 37 papers, 11 papers with code
Generative AI Beyond LLMs: System Implications of Multi-Modal Generation
As the development of large-scale Generative AI models evolve beyond text (1D) generation to include image (2D) and video (3D) generation, processing spatial and temporal information presents unique challenges to quality, performance, and efficiency.
Hardware Resilience Properties of Text-Guided Image Classifiers
This paper presents a novel method to enhance the reliability of image classification models during deployment in the face of transient hardware errors.
MAD Max Beyond Single-Node: Enabling Large Machine Learning Model Acceleration on Distributed Systems
Training and deploying large machine learning (ML) models is time-consuming and requires significant distributed computing infrastructures.
Guess & Sketch: Language Model Guided Transpilation
In this work, we leverage the strengths of LMs and symbolic solvers in a neurosymbolic approach to learned transpilation for assembly code.
INT2.1: Towards Fine-Tunable Quantized Large Language Models with Error Correction through Low-Rank Adaptation
We introduce a method that dramatically reduces fine-tuning VRAM requirements and rectifies quantization errors in quantized Large Language Models.
S$^{3}$: Increasing GPU Utilization during Generative Inference for Higher Throughput
Generating texts with a large language model (LLM) consumes massive amounts of memory.
CAMEL: Co-Designing AI Models and Embedded DRAMs for Efficient On-Device Learning
To minimize the occurrence of expensive eDRAM refresh operations, it is beneficial to shorten the lifetime of stored data during the training process.
MP-Rec: Hardware-Software Co-Design to Enable Multi-Path Recommendation
Based on our characterization of various embedding representations, we propose a hybrid embedding representation that achieves higher quality embeddings at the cost of increased memory and compute requirements.
PerfSAGE: Generalized Inference Performance Predictor for Arbitrary Deep Learning Models on Edge Devices
The ability to accurately predict deep neural network (DNN) inference performance metrics, such as latency, power, and memory footprint, for an arbitrary DNN on a target hardware platform is essential to the design of DNN based models.
GPU-based Private Information Retrieval for On-Device Machine Learning Inference
Together, for various on-device ML applications such as recommendation and language modeling, our system on a single V100 GPU can serve up to $100, 000$ queries per second -- a $>100 \times$ throughput improvement over a CPU-based baseline -- while maintaining model accuracy.
Architectural Implications of Embedding Dimension during GCN on CPU and GPU
Graph Neural Networks (GNNs) are a class of neural networks designed to extract information from the graphical structure of data.
SpeedLimit: Neural Architecture Search for Quantized Transformer Models
While research in the field of transformer models has primarily focused on enhancing performance metrics such as accuracy and perplexity, practical applications in industry often necessitate a rigorous consideration of inference latency constraints.
Tabula: Efficiently Computing Nonlinear Activation Functions for Secure Neural Network Inference
Multiparty computation approaches to secure neural network inference traditionally rely on garbled circuits for securely executing nonlinear activation functions.
Tabula: Efficiently Computing Nonlinear Activation Functions for Private Neural Network Inference
Multiparty computation approaches to private neural network inference require significant communication between server and client, incur tremendous runtime penalties, and cost massive storage overheads.
Gradient Disaggregation: Breaking Privacy in Federated Learning by Reconstructing the User Participant Matrix
We show that aggregated model updates in federated learning may be insecure.
RecPipe: Co-designing Models and Hardware to Jointly Optimize Recommendation Quality and Performance
Thus, we design RecPipeAccel (RPAccel), a custom accelerator that jointly optimizes quality, tail-latency, and system throughput.
Quantifying and Maximizing the Benefits of Back-End Noise Adaption on Attention-Based Speech Recognition Models
This work analyzes how attention-based Bidirectional Long Short-Term Memory (BLSTM) models adapt to noise-augmented speech.
AutoPilot: Automating SoC Design Space Exploration for SWaP Constrained Autonomous UAVs
Balancing a computing system for a UAV requires considering both the cyber (e. g., sensor rate, compute performance) and physical (e. g., payload weight) characteristics that affect overall performance.
RecSSD: Near Data Processing for Solid State Drive Based Recommendation Inference
Neural personalized recommendation models are used across a wide variety of datacenter applications including search, social media, and entertainment.
EdgeBERT: Sentence-Level Energy Optimizations for Latency-Aware Multi-Task NLP Inference
Transformer-based language models such as BERT provide significant accuracy improvement for a multitude of natural language processing (NLP) tasks.
Cross-Stack Workload Characterization of Deep Recommendation Systems
Deep learning based recommendation systems form the backbone of most personalized cloud services.
CHIPKIT: An agile, reusable open-source framework for rapid test chip development
The current trend for domain-specific architectures (DSAs) has led to renewed interest in research test chips to demonstrate new specialized hardware.
Hardware Architecture
DeepRecSys: A System for Optimizing End-To-End At-scale Neural Recommendation Inference
Neural personalized recommendation is the corner-stone of a wide collection of cloud services and products, constituting significant compute demand of the cloud infrastructure.
Distributed, Parallel, and Cluster Computing
SMAUG: End-to-End Full-Stack Simulation Infrastructure for Deep Learning Workloads
In recent years, there has been tremendous advances in hardware acceleration of deep neural networks.
A binary-activation, multi-level weight RNN and training algorithm for ADC-/DAC-free and noise-resilient processing-in-memory inference with eNVM
We propose a new algorithm for training neural networks with binary activations and multi-level weights, which enables efficient processing-in-memory circuits with embedded nonvolatile memories (eNVM).
MLPerf Training Benchmark
Machine learning (ML) needs industry-standard performance benchmarks to support design and competitive evaluation of the many emerging software and hardware solutions for ML.
AdaptivFloat: A Floating-point based Data Type for Resilient Deep Learning Inference
Conventional hardware-friendly quantization methods, such as fixed-point or integer, tend to perform poorly at very low word sizes as their shrinking dynamic ranges cannot adequately capture the wide data distributions commonly seen in sequence transduction models.
Decoupling Weight Regularization from Batch Size for Model Compression
Using various models, we show that simple weight updates to comply with compression formats along with long NR period is enough to achieve high compression ratio and model accuracy.
Network Pruning for Low-Rank Binary Index
In this paper, we propose a new network pruning technique that generates a low-rank binary index matrix to compress index data significantly.
MASR: A Modular Accelerator for Sparse RNNs
The architecture is enhanced by a series of dynamic activation optimizations that enable compact storage, ensure no energy is wasted computing null operations, and maintain high MAC utilization for highly parallel accelerator designs.
Benchmarking TPU, GPU, and CPU Platforms for Deep Learning
Training deep learning models is compute-intensive and there is an industry-wide trend towards hardware specialization to improve performance.
Learning Low-Rank Approximation for CNNs
Low-rank approximation is an effective model compression technique to not only reduce parameter storage requirements, but to also reduce computations.
Structured Compression by Weight Encryption for Unstructured Pruning and Quantization
Model compression techniques, such as pruning and quantization, are becoming increasingly important to reduce the memory footprints and the amount of computations.
Network Pruning for Low-Rank Binary Indexing
Pruning is an efficient model compression technique to remove redundancy in the connectivity of deep neural networks (DNNs).
Cloud No Longer a Silver Bullet, Edge to the Rescue
This paper takes the position that, while cognitive computing today relies heavily on the cloud, we will soon see a paradigm shift where cognitive computing primarily happens on network edges.
Weightless: Lossy Weight Encoding For Deep Neural Network Compression
This results in up to a 1. 51x improvement over the state-of-the-art.
Fathom: Reference Workloads for Modern Deep Learning Methods
Fathom has been released online, and this paper focuses on understanding the fundamental performance characteristics of each model.
