{"title":"通过分支和边界缩小动态范围","authors":"Thore Gerlach, Nico Piatkowski","doi":"arxiv-2409.10863","DOIUrl":null,"url":null,"abstract":"The demand for high-performance computing in machine learning and artificial\nintelligence has led to the development of specialized hardware accelerators\nlike Tensor Processing Units (TPUs), Graphics Processing Units (GPUs), and\nField-Programmable Gate Arrays (FPGAs). A key strategy to enhance these\naccelerators is the reduction of precision in arithmetic operations, which\nincreases processing speed and lowers latency - crucial for real-time AI\napplications. Precision reduction minimizes memory bandwidth requirements and\nenergy consumption, essential for large-scale and mobile deployments, and\nincreases throughput by enabling more parallel operations per cycle, maximizing\nhardware resource utilization. This strategy is equally vital for solving\nNP-hard quadratic unconstrained binary optimization (QUBO) problems common in\nmachine learning, which often require high precision for accurate\nrepresentation. Special hardware solvers, such as quantum annealers, benefit\nsignificantly from precision reduction. This paper introduces a fully\nprincipled Branch-and-Bound algorithm for reducing precision needs in QUBO\nproblems by utilizing dynamic range as a measure of complexity. Experiments\nvalidate our algorithm's effectiveness on an actual quantum annealer.","PeriodicalId":501286,"journal":{"name":"arXiv - MATH - Optimization and Control","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2024-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Dynamic Range Reduction via Branch-and-Bound\",\"authors\":\"Thore Gerlach, Nico Piatkowski\",\"doi\":\"arxiv-2409.10863\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"The demand for high-performance computing in machine learning and artificial\\nintelligence has led to the development of specialized hardware accelerators\\nlike Tensor Processing Units (TPUs), Graphics Processing Units (GPUs), and\\nField-Programmable Gate Arrays (FPGAs). A key strategy to enhance these\\naccelerators is the reduction of precision in arithmetic operations, which\\nincreases processing speed and lowers latency - crucial for real-time AI\\napplications. Precision reduction minimizes memory bandwidth requirements and\\nenergy consumption, essential for large-scale and mobile deployments, and\\nincreases throughput by enabling more parallel operations per cycle, maximizing\\nhardware resource utilization. This strategy is equally vital for solving\\nNP-hard quadratic unconstrained binary optimization (QUBO) problems common in\\nmachine learning, which often require high precision for accurate\\nrepresentation. Special hardware solvers, such as quantum annealers, benefit\\nsignificantly from precision reduction. This paper introduces a fully\\nprincipled Branch-and-Bound algorithm for reducing precision needs in QUBO\\nproblems by utilizing dynamic range as a measure of complexity. Experiments\\nvalidate our algorithm's effectiveness on an actual quantum annealer.\",\"PeriodicalId\":501286,\"journal\":{\"name\":\"arXiv - MATH - Optimization and Control\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-17\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - MATH - Optimization and Control\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.10863\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - MATH - Optimization and Control","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.10863","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
The demand for high-performance computing in machine learning and artificial
intelligence has led to the development of specialized hardware accelerators
like Tensor Processing Units (TPUs), Graphics Processing Units (GPUs), and
Field-Programmable Gate Arrays (FPGAs). A key strategy to enhance these
accelerators is the reduction of precision in arithmetic operations, which
increases processing speed and lowers latency - crucial for real-time AI
applications. Precision reduction minimizes memory bandwidth requirements and
energy consumption, essential for large-scale and mobile deployments, and
increases throughput by enabling more parallel operations per cycle, maximizing
hardware resource utilization. This strategy is equally vital for solving
NP-hard quadratic unconstrained binary optimization (QUBO) problems common in
machine learning, which often require high precision for accurate
representation. Special hardware solvers, such as quantum annealers, benefit
significantly from precision reduction. This paper introduces a fully
principled Branch-and-Bound algorithm for reducing precision needs in QUBO
problems by utilizing dynamic range as a measure of complexity. Experiments
validate our algorithm's effectiveness on an actual quantum annealer.