arXiv subjects

Vijay Gadepally

Publications and source records attributed to Vijay Gadepally.

At least 37 records · Page 2Linked to original sources

arXiv2024

Sustainable Supercomputing for AI: GPU Power Capping at HPC Scale

As research and deployment of AI grows, the computational burden to support and sustain its progress inevitably does too. To train or fine-tune state-of-the-art models in NLP, computer vision, etc., some form of AI hardware acceleration is virtually a requirement. Recent large language models require considerable resources to train and deploy, resulting in significant energy usage, potential carbon emissions, and massive demand for GPUs and other hardware accelerators. However, this surge carries large implications for energy sustainability at the HPC/datacenter level. In this paper, we study the aggregate effect of power-capping GPUs on GPU temperature and power draw at a research supercomputing center. With the right amount of power-capping, we show significant decreases in both temperature and power draw, reducing power consumption and potentially improving hardware life-span with minimal impact on job performance. While power-capping reduces power draw by design, the aggregate system-wide effect on overall energy consumption is less clear; for instance, if users notice job performance degradation from GPU power-caps, they may request additional GPU-jobs to compensate, negating any energy savings or even worsening energy consumption. To our knowledge, our work is the first to conduct and make available a detailed analysis of the effects of GPU power-capping at the supercomputing scale. We hope our work will inspire HPCs/datacenters to further explore, evaluate, and communicate the impact of power-capping AI hardware accelerators for more sustainable AI.

cs.AR↗

arXiv2023

Deployment of Real-Time Network Traffic Analysis using GraphBLAS Hypersparse Matrices and D4M Associative Arrays

Matrix/array analysis of networks can provide significant insight into their behavior and aid in their operation and protection. Prior work has demonstrated the analytic, performance, and compression capabilities of GraphBLAS (graphblas.org) hypersparse matrices and D4M (d4m.mit.edu) associative arrays (a mathematical superset of matrices). Obtaining the benefits of these capabilities requires integrating them into operational systems, which comes with its own unique challenges. This paper describes two examples of real-time operational implementations. First, is an operational GraphBLAS implementation that constructs anonymized hypersparse matrices on a high-bandwidth network tap. Second, is an operational D4M implementation that analyzes daily cloud gateway logs. The architectures of these implementations are presented. Detailed measurements of the resources and the performance are collected and analyzed. The implementations are capable of meeting their operational requirements using modest computational resources (a couple of processing cores). GraphBLAS is well-suited for low-level analysis of high-bandwidth connections with relatively structured network data. D4M is well-suited for higher-level analysis of more unstructured data. This work demonstrates that these technologies can be implemented in operational settings.

cs.NI↗

arXiv2023

Toward Sustainable HPC: Carbon Footprint Estimation and Environmental Implications of HPC Systems

The rapid growth in demand for HPC systems has led to a rise in carbon footprint, which requires urgent intervention. In this work, we present a comprehensive analysis of the carbon footprint of high-performance computing (HPC) systems, considering the carbon footprint during both the hardware manufacturing and system operational stages. Our work employs HPC hardware component carbon footprint modeling, regional carbon intensity analysis, and experimental characterization of the system life cycle to highlight the importance of quantifying the carbon footprint of HPC systems.

cs.DC↗

arXiv2023

Lincoln AI Computing Survey (LAICS) Update

This paper is an update of the survey of AI accelerators and processors from past four years, which is now called the Lincoln AI Computing Survey - LAICS (pronounced "lace"). As in past years, this paper collects and summarizes the current commercial accelerators that have been publicly announced with peak performance and peak power consumption numbers. The performance and power values are plotted on a scatter graph, and a number of dimensions and observations from the trends on this plot are again discussed and analyzed. Market segments are highlighted on the scatter plot, and zoomed plots of each segment are also included. Finally, a brief description of each of the new accelerators that have been added in the survey this year is included.

cs.AI↗

arXiv2023

From Words to Watts: Benchmarking the Energy Costs of Large Language Model Inference

Large language models (LLMs) have exploded in popularity due to their new generative capabilities that go far beyond prior state-of-the-art. These technologies are increasingly being leveraged in various domains such as law, finance, and medicine. However, these models carry significant computational challenges, especially the compute and energy costs required for inference. Inference energy costs already receive less attention than the energy costs of training LLMs -- despite how often these large models are called on to conduct inference in reality (e.g., ChatGPT). As these state-of-the-art LLMs see increasing usage and deployment in various domains, a better understanding of their resource utilization is crucial for cost-savings, scaling performance, efficient hardware usage, and optimal inference strategies. In this paper, we describe experiments conducted to study the computational and energy utilization of inference with LLMs. We benchmark and conduct a preliminary analysis of the inference performance and inference energy costs of different sizes of LLaMA -- a recent state-of-the-art LLM -- developed by Meta AI on two generations of popular GPUs (NVIDIA V100 \& A100) and two datasets (Alpaca and GSM8K) to reflect the diverse set of tasks/benchmarks for LLMs in research and practice. We present the results of multi-node, multi-GPU inference using model sharding across up to 32 GPUs. To our knowledge, our work is the one of the first to study LLM inference performance from the perspective of computational and energy resources at this scale.

cs.CL↗

arXiv2023

Mapping of Internet "Coastlines" via Large Scale Anonymized Network Source Correlations

Expanding the scientific tools available to protect computer networks can be aided by a deeper understanding of the underlying statistical distributions of network traffic and their potential geometric interpretations. Analyses of large scale network observations provide a unique window into studying those underlying statistics. Newly developed GraphBLAS hypersparse matrices and D4M associative array technologies enable the efficient anonymized analysis of network traffic on the scale of trillions of events. This work analyzes over 100,000,000,000 anonymized packets from the largest Internet telescope (CAIDA) and over 10,000,000 anonymized sources from the largest commercial honeyfarm (GreyNoise). Neither CAIDA nor GreyNoise actively emit Internet traffic and provide distinct observations of unsolicited Internet traffic (primarily botnets and scanners). Analysis of these observations confirms the previously observed Cauchy-like distributions describing temporal correlations between Internet sources. The Gull lighthouse problem is a well-known geometric characterization of the standard Cauchy distribution and motivates a potential geometric interpretation for Internet observations. This work generalizes the Gull lighthouse problem to accommodate larger classes of coastlines, deriving a closed-form solution for the resulting probability distributions, stating and examining the inverse problem of identifying an appropriate coastline given a continuous probability distribution, identifying a geometric heuristic for solving this problem computationally, and applying that heuristic to examine the temporal geometry of different subsets of network observations. Application of this method to the CAIDA and GreyNoise data reveals a several orders of magnitude difference between known benign and other traffic which can lead to potentially novel ways to protect networks.

cs.SI↗

arXiv2023

SeeSaw: Interactive Ad-hoc Search Over Image Databases

As image datasets become ubiquitous, the problem of ad-hoc searches over image data is increasingly important. Many high-level data tasks in machine learning, such as constructing datasets for training and testing object detectors, imply finding ad-hoc objects or scenes within large image datasets as a key sub-problem. New foundational visual-semantic embeddings trained on massive web datasets such as Contrastive Language-Image Pre-Training (CLIP) can help users start searches on their own data, but we find there is a long tail of queries where these models fall short in practice. SeeSaw is a system for interactive ad-hoc searches on image datasets that integrates state-of-the-art embeddings like CLIP with user feedback in the form of box annotations to help users quickly locate images of interest in their data even in the long tail of harder queries. One key challenge for SeeSaw is that, in practice, many sensible approaches to incorporating feedback into future results, including state-of-the-art active-learning algorithms, can worsen results compared to introducing no feedback, partly due to CLIP's high-average performance. Therefore, SeeSaw includes several algorithms that empirically result in larger and also more consistent improvements. We compare SeeSaw's accuracy to both using CLIP alone and to a state-of-the-art active-learning baseline and find SeeSaw consistently helps improve results for users across four datasets and more than a thousand queries. SeeSaw increases Average Precision (AP) on search tasks by an average of .08 on a wide benchmark (from a base of .72), and by a .27 on a subset of more difficult queries where CLIP alone performs poorly.

cs.DB↗

arXiv2023

pPython Performance Study

pPython seeks to provide a parallel capability that provides good speed-up without sacrificing the ease of programming in Python by implementing partitioned global array semantics (PGAS) on top of a simple file-based messaging library (PythonMPI) in pure Python. pPython follows a SPMD (single program multiple data) model of computation. pPython runs on a single-node (e.g., a laptop) running Windows, Linux, or MacOS operating systems or on any combination of heterogeneous systems that support Python, including on a cluster through a Slurm scheduler interface so that pPython can be executed in a massively parallel computing environment. It is interesting to see what performance pPython can achieve compared to the traditional socket-based MPI communication because of its unique file-based messaging implementation. In this paper, we present the point-to-point and collective communication performances of pPython and compare them with those obtained by using mpi4py with OpenMPI. For large messages, pPython demonstrates comparable performance as compared to mpi4py.

cs.DC↗

arXiv2023

Focusing and Calibration of Large Scale Network Sensors using GraphBLAS Anonymized Hypersparse Matrices

Defending community-owned cyber space requires community-based efforts. Large-scale network observations that uphold the highest regard for privacy are key to protecting our shared cyberspace. Deployment of the necessary network sensors requires careful sensor placement, focusing, and calibration with significant volumes of network observations. This paper demonstrates novel focusing and calibration procedures on a multi-billion packet dataset using high-performance GraphBLAS anonymized hypersparse matrices. The run-time performance on a real-world data set confirms previously observed real-time processing rates for high-bandwidth links while achieving significant data compression. The output of the analysis demonstrates the effectiveness of these procedures at focusing the traffic matrix and revealing the underlying stable heavy-tail statistical distributions that are necessary for anomaly detection. A simple model of the corresponding probability of detection ($p_{\rm d}$) and probability of false alarm ($p_{\rm fa}$) for these distributions highlights the criticality of network sensor focusing and calibration. Once a sensor is properly focused and calibrated it is then in a position to carry out two of the central tenets of good cybersecurity: (1) continuous observation of the network and (2) minimizing unbrokered network connections.

cs.NI↗

arXiv2023

Clover: Toward Sustainable AI with Carbon-Aware Machine Learning Inference Service

This paper presents a solution to the challenge of mitigating carbon emissions from hosting large-scale machine learning (ML) inference services. ML inference is critical to modern technology products, but it is also a significant contributor to carbon footprint. We introduce Clover, a carbon-friendly ML inference service runtime system that balances performance, accuracy, and carbon emissions through mixed-quality models and GPU resource partitioning. Our experimental results demonstrate that Clover is effective in substantially reducing carbon emissions while maintaining high accuracy and meeting service level agreement (SLA) targets.

cs.DC↗

arXiv2023

KAIROS: Building Cost-Efficient Machine Learning Inference Systems with Heterogeneous Cloud Resources

Online inference is becoming a key service product for many businesses, deployed in cloud platforms to meet customer demands. Despite their revenue-generation capability, these services need to operate under tight Quality-of-Service (QoS) and cost budget constraints. This paper introduces KAIROS, a novel runtime framework that maximizes the query throughput while meeting QoS target and a cost budget. KAIROS designs and implements novel techniques to build a pool of heterogeneous compute hardware without online exploration overhead, and distribute inference queries optimally at runtime. Our evaluation using industry-grade deep learning (DL) models shows that KAIROS yields up to 2X the throughput of an optimal homogeneous solution, and outperforms state-of-the-art schemes by up to 70%, despite advantageous implementations of the competing schemes to ignore their exploration overhead.

cs.DC↗

arXiv2023

A Green(er) World for A.I

As research and practice in artificial intelligence (A.I.) grow in leaps and bounds, the resources necessary to sustain and support their operations also grow at an increasing pace. While innovations and applications from A.I. have brought significant advances, from applications to vision and natural language to improvements to fields like medical imaging and materials engineering, their costs should not be neglected. As we embrace a world with ever-increasing amounts of data as well as research and development of A.I. applications, we are sure to face an ever-mounting energy footprint to sustain these computational budgets, data storage needs, and more. But, is this sustainable and, more importantly, what kind of setting is best positioned to nurture such sustainable A.I. in both research and practice? In this paper, we outline our outlook for Green A.I. -- a more sustainable, energy-efficient and energy-aware ecosystem for developing A.I. across the research, computing, and practitioner communities alike -- and the steps required to arrive there. We present a bird's eye view of various areas for potential changes and improvements from the ground floor of AI's operational and hardware optimizations for datacenters/HPCs to the current incentive structures in the world of A.I. research and practice, and more. We hope these points will spur further discussion, and action, on some of these issues and their potential solutions.

cs.AI↗

arXiv2022

Python Implementation of the Dynamic Distributed Dimensional Data Model

Python has become a standard scientific computing language with fast-growing support of machine learning and data analysis modules, as well as an increasing usage of big data. The Dynamic Distributed Dimensional Data Model (D4M) offers a highly composable, unified data model with strong performance built to handle big data fast and efficiently. In this work we present an implementation of D4M in Python. $D4M.py$ implements all foundational functionality of D4M and includes Accumulo and SQL database support via Graphulo. We describe the mathematical background and motivation, an explanation of the approaches made for its fundamental functions and building blocks, and performance results which compare $D4M.py$'s performance to D4M-MATLAB and D4M.jl.

cs.DC↗

arXiv2022

AI and ML Accelerator Survey and Trends

This paper updates the survey of AI accelerators and processors from past three years. This paper collects and summarizes the current commercial accelerators that have been publicly announced with peak performance and power consumption numbers. The performance and power values are plotted on a scatter graph, and a number of dimensions and observations from the trends on this plot are again discussed and analyzed. Two new trends plots based on accelerator release dates are included in this year's paper, along with the additional trends of some neuromorphic, photonic, and memristor-based inference accelerators.

cs.AR↗

arXiv2022

MISO: Exploiting Multi-Instance GPU Capability on Multi-Tenant Systems for Machine Learning

GPU technology has been improving at an expedited pace in terms of size and performance, empowering HPC and AI/ML researchers to advance the scientific discovery process. However, this also leads to inefficient resource usage, as most GPU workloads, including complicated AI/ML models, are not able to utilize the GPU resources to their fullest extent -- encouraging support for GPU multi-tenancy. We propose MISO, a technique to exploit the Multi-Instance GPU (MIG) capability on the latest NVIDIA datacenter GPUs (e.g., A100, H100) to dynamically partition GPU resources among co-located jobs. MISO's key insight is to use the lightweight, more flexible Multi-Process Service (MPS) capability to predict the best MIG partition allocation for different jobs, without incurring the overhead of implementing them during exploration. Due to its ability to utilize GPU resources more efficiently, MISO achieves 49% and 16% lower average job completion time than the unpartitioned and optimal static GPU partition schemes, respectively.

cs.DC↗

arXiv2022

Hypersparse Network Flow Analysis of Packets with GraphBLAS

Internet analysis is a major challenge due to the volume and rate of network traffic. In lieu of analyzing traffic as raw packets, network analysts often rely on compressed network flows (netflows) that contain the start time, stop time, source, destination, and number of packets in each direction. However, many traffic analyses benefit from temporal aggregation of multiple simultaneous netflows, which can be computationally challenging. To alleviate this concern, a novel netflow compression and resampling method has been developed leveraging GraphBLAS hyperspace traffic matrices that preserve anonymization while enabling subrange analysis. Standard multitemporal spatial analyses are then performed on each subrange to generate detailed statistical aggregates of the source packets, source fan-out, unique links, destination fan-in, and destination packets of each subrange which can then be used for background modeling and anomaly detection. A simple file format based on GraphBLAS sparse matrices is developed for storing these statistical aggregates. This method is scale tested on the MIT SuperCloud using a 50 trillion packet netflow corpus from several hundred sites collected over several months. The resulting compression achieved is significant (<0.1 bit per packet) enabling extremely large netflow analyses to be stored and transported. The single node parallel performance is analyzed in terms of both processors and threads showing that a single node can perform hundreds of simultaneous analyses at over a million packets/sec (roughly equivalent to a 10 Gigabit link).

cs.NI↗

arXiv2022

An Evaluation of Low Overhead Time Series Preprocessing Techniques for Downstream Machine Learning

In this paper we address the application of pre-processing techniques to multi-channel time series data with varying lengths, which we refer to as the alignment problem, for downstream machine learning. The misalignment of multi-channel time series data may occur for a variety of reasons, such as missing data, varying sampling rates, or inconsistent collection times. We consider multi-channel time series data collected from the MIT SuperCloud High Performance Computing (HPC) center, where different job start times and varying run times of HPC jobs result in misaligned data. This misalignment makes it challenging to build AI/ML approaches for tasks such as compute workload classification. Building on previous supervised classification work with the MIT SuperCloud Dataset, we address the alignment problem via three broad, low overhead approaches: sampling a fixed subset from a full time series, performing summary statistics on a full time series, and sampling a subset of coefficients from time series mapped to the frequency domain. Our best performing models achieve a classification accuracy greater than 95%, outperforming previous approaches to multi-channel time series classification with the MIT SuperCloud Dataset by 5%. These results indicate our low overhead approaches to solving the alignment problem, in conjunction with standard machine learning techniques, are able to achieve high levels of classification accuracy, and serve as a baseline for future approaches to addressing the alignment problem, such as kernel methods.

cs.LG↗

arXiv2022

GraphBLAS on the Edge: Anonymized High Performance Streaming of Network Traffic

Long range detection is a cornerstone of defense in many operating domains (land, sea, undersea, air, space, ..,). In the cyber domain, long range detection requires the analysis of significant network traffic from a variety of observatories and outposts. Construction of anonymized hypersparse traffic matrices on edge network devices can be a key enabler by providing significant data compression in a rapidly analyzable format that protects privacy. GraphBLAS is ideally suited for both constructing and analyzing anonymized hypersparse traffic matrices. The performance of GraphBLAS on an Accolade Technologies edge network device is demonstrated on a near worse case traffic scenario using a continuous stream of CAIDA Telescope darknet packets. The performance for varying numbers of traffic buffers, threads, and processor cores is explored. Anonymized hypersparse traffic matrices can be constructed at a rate of over 50,000,000 packets per second; exceeding a typical 400 Gigabit network link. This performance demonstrates that anonymized hypersparse traffic matrices are readily computable on edge network devices with minimal compute resources and can be a viable data product for such devices.

cs.NI↗