The Case for Energy-Proportional Computing

By Luiz André Barroso, Urs Hölzle. 2007

2,662 citations recorded by the source corpus.

Subfields: Computer Networks and Communications, Hardware and Architecture, Information Systems, ML Systems & Hardware

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

In Hardware and Architecture

In Information Systems

In ML Systems & Hardware

Citations in the corpus

In-datacenter performance analysis of a tensor processing unit (2017, methodology): “ower, as you must supply sufficient power and cooling when hardware is at full power. However, the cost of electricity is based on the a v e r age consumed as the workload varies during the day. [Bar07] found that servers are 100% busy less than 10% of the time and advocated e ne r gy pr opor t i ”
A Survey of Green Networking Research (2010, methodology): “Therefore, as highlighted in [17], most mobile devices and wireless sensors tend to have a low duty cycle, operating by means of short periods of intense performance, followed by long idle intervals.”
Toward Edge Intelligence: Multiaccess Edge Computing for 5G and Internet of Things (2020, background): “The energy-proportional server means a server whose energy consumption is in proportion to its computation workload [159].”
A survey on techniques for improving the energy efficiency of large-scale distributed systems (2014, background): “In addition to improving the energy-eﬃciency of hardware components, to get closer to proportional computing [Barroso and Holzle 2007], manufacturers could allow individual components such as PCI slots and CPU cores, to be put into sleep state when not in use (independently from the motherboard).”
AutoScale: Dynamic, Robust Capacity Management for Multi-Tier Data Centers (2012, methodology): “We introduce a dynamic capacity management policy, AutoScale, that greatly reduces the number of servers needed in data centers driven by unpredictable, time-varying load, while meeting response time SLAs.”
A Power Benchmarking Framework for Network Devices (2009, background): “– Ideally, devices should consume energy proportional to their load [6, 7].”
IX: A Protected Dataplane Operating System for High Throughput and Low Latency (2014, background): “Our work focuses on improving operating systems for applications with aggressive networking requirements running on multi-core servers.”
Minimizing Electricity Cost: Optimization of Distributed Internet Data Centers in a Multi-Electricity-Market Environment (2010, background): “The adoption of low power chipset with chip multiprocessing [ 13 ], virtualization technologies [24], [21], power control theories [34], and other energy saving solutions [23], [14] have also contributed positively toward more energy-efficient servers.”
A Survey on Mobile Edge Computing: The Communication Perspective (2017, background): “, the energy consumption of a server should be proportional to its computation load [190].”
GreenCloud: a new architecture for green data center (2009, methodology): “In [25], Barrsos et al, studied how to use Chip Multi-Processor (CMP) to achieve power management.”
PowerNap: eliminating server idle power (2009, background): “At idle, current servers still draw about 60% of peak power [1, 6, 13].”
GreenSlot: Scheduling energy consumption in green datacenters (2011, background): “This approach is ideal in terms of conserving energy, since keeping modern servers powered on involves a high “static” energy [2].”
Capping the brown energy consumption of Internet services at low cost (2010, background): “By default, we study servers that are perfectly energy-proportional [3], i.”