Every Joule is Precious Carla Schlatter Ellis Duke University

1
Every Joule is PreciousCarla Schlatter
EllisDuke University
Milly Watt Project
Systems Architecture
2
Energy in Computing

• Energy for computing is an important problem(
  not just for mobile computing)
• Reducing heat production and fan noise
• Extending battery life for mobile/wireless
  devices
• Conserving energy resources (lessen environmental
  impact, save on electricity costs)
• Energy should be a first class resource at
  upper levels of system design

3
Understanding the Energy Problem

• Energy (Joules) Power (watts) Time (sec)
• E P t
• Power (watts) Voltage (volts) Current (amps)
• P V I
• Current (amps) Voltage (volts) / Resistance
  (ohms)
• I V / R

4
Demand SideHW Power Budget
CPU
Cache
Memory Bus
I/O Bridge
I/O Bus
Main Memory
Disk Controller
Graphics Controller
Network Interface
Graphics
Disk
Disk
Network
Intel targets
5
Reducing Demand
HW / SW Cooperation

• Software
• High level
• Coarse grain
• OS, compiler or application
• Affect usage patterns

Hardware

• Voltage Scaling
• Clock gating
• Power modes Turning off HW blocks

• Low level
• Fine grain
• Low-power Circuits

• Re-examine interactions between HW and SW,
  particularly within the resource management
  functions of the Operating System

6
Supply Side Battery Properties

• Battery models provide strategyBattery lifetime
  can be determined by controlling discharge
  rate.Limiting availability of currentcy.

7
Battery Discharge Behavior
Discharge behavior of lithium-ion cell withVoc
3V and Vcut 1V
8
Energy Goals /Metrics

• Battery lifetime (hours)
• Energy usage (by fixed task set) (Joules)
• Energy Delay (penalizes achieving energy
  saving by bad performance)
• Work units / joule (e.g. Mbytes/joule or
  MIPS/joule)
• Work / battery discharge.
• Thermal limits (constrained power) (Watts)

9
How to Reduce Energy Consumption?

• Energy S Poweri x Timei
• To reduce energy used for task
• Reduce power cost of power state ithrough better
  technology.
• Reduce time spent in the higher cost power
  states.
• Amortize transition states, if significant.

i e powerstates
10
Energy the OS

• Traditionally, the system-wide view of resources
  and workload demands resides with the OS
• Explicitly managing energy will require
  coordination with typical resource management
• Energy is not just another resource
• Energy has a impact on every other resource of a
  computing system it is central.
• A focus on energy provides an opportunity to
  rethink OS design

11
Traditional Influences in OS Design
Scientific computationsDatabase operations
Multi-user
Workload
Services API
Goals/Metrics
Internal Structure
Policies / Mechanisms
Performance asBandwidth and Latency.
Hardware Resources
Processor, Memory, Disks, Network
12
Rethinking OS Design

• What is the impact of changing the primary
  goal of the OS to energy rather than
  (speed-based) performance?
• Affects every aspect of OS services and
  structure
• Interfaces needed by applications that want to
  affect power consumption
• Internal organization and algorithms
• Resource management policies and mechanisms

13
Rethinking OS Design
Productivity applications, Games,
Multimedia, Web access,
Personal (PDAs),
Embedded, E-Commerce.
Workload
Services API
Goals/Metrics
Internal Structure
Policies / Mechanisms
Energy
Hardware Resources
Processor, Memory, Disks, Wireless
networking, Mic Speaker, Motors Sensors,
Batteries
14
Rethinking OS Design
Non-energy-awaregeneral purpose applications
Workload
Services API
Goals/Metrics
Internal Structure
Policies / Mechanisms
Batterylifetime
Hardware Resources
Battery-powered Laptop
15
Related Work

• Energy-unaware OS
• Low-power hardware, energy-aware compilers,
  algorithm development
• Services (Chase MUSE)
• Energy-aware OS with Unaware applications
• Per-device solutions (disk spindown, DVS)
• Energy-aware OS with cooperating Energy-aware
  Applications
• Flinn Odyssey (fidelity-based), Bellosa Coop
  I/O, Nemesis OS

16
Milly Watt Activities

• ECOSystem Explicitly managing energy via the OS
  (ASPLOS02, USENIX03)
• Power-aware memory(ASPLOS00, ISLPED01, PACS02,
  PACS03)
• FaceOff Sensor-based display power management
  (HOTOS03, Mobisys Context Aware 04)

17
Milly Watt Activities

• ECOSystem Explicitly managing energy via the OS
  (ASPLOS02, USENIX03)
• Power-aware memory(ASPLOS00, ISLPED01, PACS02,
  PACS03)
• FaceOff Sensor-based display power management
  (HOTOS03, Mobisys Context Aware 04)

18
Milly Watt Activities

• ECOSystem Explicitly managing energy via the OS
  (ASPLOS02, USENIX03)
• Power-aware memory(ASPLOS00, ISLPED01, PACS02,
  PACS03)
• FaceOff Sensor-based display power management
  (HOTOS03, Mobisys Context Aware 04)

19
For More Information

• www.cs.duke.edu/ari/millywatt/
• email carla_at_cs.duke.edu
• Experiences in Managing Energy with
  ECOSystem,Heng Zeng, Carla Ellis, and Alvin
  Lebeck, IEEE Pervasive Computing, January-March
  2005.
• Currentcy Unifying Policies for Resource
  Management,H. Zeng, C. Ellis, A. Lebeck, A.
  Vahdat, in USENIX 2003 Annual Technical
  Conference
• ECOSystem Managing Energy as a First Class
  Operating System Resource, H. Zeng, X. Fan, C.
  Ellis, A. Lebeck, and A. Vahdat, Proceedings of
  ASPLOS 2002