课题基金 / 基金详情

Physically Aware Computer Architecture

Physically Aware Computer Architecture
物理感知计算机架构
批准号:
0429765
负责人:
Kevin Skadron
金额:
$0.0万
依托单位国家:
美国
项目类别:
Continuing Grant
财政年份:
2004
资助国家:
美国
项目状态:
已结题
起止时间:
2004-09-15 至 2007-08-31

项目摘要

项目成果

Kevin Skadron的其他基金

相似基金

相关文献

中文摘要
翻译
物理感知计算机体系结构摘要技术扩展和低级别物理影响使抽象级别之间的干净接口变得紧张,这是使设计过程可管理所必需的。随着新的物理现象在基于纳米技术的电路中显现,这种张力将会增加。与此同时,微体系结构在管理物理现象对性能、能源效率、成本和预期寿命等关键指标的影响方面正变得越来越重要。该方案考虑了计算机体系结构领域和物理领域方面之间的相互作用,如热效应、老化过程、布局规划、互连、工艺变化和电源变化。我们建议使用“紧凑”模型来模拟这些现象。我们将开发关键物理现象的模型,并探索管理其影响的体系结构技术。除非微体系结构可以解释工艺、电压和温度变化等低级别物理现象,否则增加设计余量将带来高昂的、甚至可能是令人望而却步的成本,从而极大地限制设计的可能性。对这些现象的建模和设计方法的相似性的发现提供了一个解决方案。在更广泛的影响方面,该项目将开发一个灵活的建模基础设施,使架构和计算机系统设计界能够执行物理感知设计和分析,并弥补这些技术趋势。进一步的影响将出现在教育一代学生考虑更高层次的抽象设计领域中的低层次物理现象。
英文摘要
Physically Aware Computer ArchitectureAbstractTechnology scaling and low-level physical effects are straining the clean interfaces between abstraction levels necessary to make the design process manageable. This tension will increase as new physical phenomena are manifested in nanotechnology-based circuits. At the same time, microarchitecture is becoming increasingly important in managing the way physical phenomena impact key figures of merit such as performance, energy efficiency, cost, and expected lifetime. This proposal considers the interaction between the computer architecture domain and physical domain aspects such as thermal effects, aging processes, floorplanning, interconnect, process variations, and power-supply variation. We propose to model these phenomena using "compact" models. We will develop models for key physical phenomena and explore architecture techniques for managing their impact.Unless microarchitecture can account for low-level physical phenomena like process, voltage, and temperature variability, increasing design margins will impose high and possibly prohibitive costs that drastically limit design possibilities. The discovery of similarities in the modeling and design approaches for such phenomena offers a solution. In terms of broader impact, the project will develop a flexible modeling infrastructure that will enable the architecture and computer systems design community to perform physically-aware design and analysis and compensate for these technology trends. Further impact will arise from educating a generation of students to consider low-level physical phenomena in higher-level abstract design domains.
期刊论文(0)
专著(0)
科研奖励(0)
会议论文
Collaborative Research: SHF: Medium: A hardware-software co-design approach for high-performance in-memory analytic data processing
  • 批准号:
    2312740
  • 项目类别:
    Standard Grant
  • 资助金额:
    $40.0万
  • 财政年份:
    2023
  • 负责人:
    Kevin Skadron
  • 依托单位:
CI-New: Community Infrastructure Supporting Hardware Acceleration Research and Education
  • 批准号:
    1730606
  • 项目类别:
    Standard Grant
  • 资助金额:
    $49.96万
  • 财政年份:
    2017
  • 负责人:
    Kevin Skadron
  • 依托单位:
XPS:FULL: New Abstractions and Applications for Automata Computing
  • 批准号:
    1629450
  • 项目类别:
    Standard Grant
  • 资助金额:
    $87.5万
  • 财政年份:
    2016
  • 负责人:
    Kevin Skadron
  • 依托单位:
CI-P: Community Infrastructure to Catalyze Research in Automata Computing
  • 批准号:
    1513188
  • 项目类别:
    Standard Grant
  • 资助金额:
    $9.98万
  • 财政年份:
    2015
  • 负责人:
    Kevin Skadron
  • 依托单位:
海外基金