Leveraging Renewable Energy in Data Centers: Present and Future - - PowerPoint PPT Presentation

▶
leveraging renewable energy in data centers present and
SMART_READER_LITE
LIVE PREVIEW

Leveraging Renewable Energy in Data Centers: Present and Future - - PowerPoint PPT Presentation

Jun 25, 2023 185 likes •571 views

Leveraging Renewable Energy in Data Centers: Present and Future Ricardo Bianchini Department of Computer Science Collaborators: Josep L. Berral, Inigo Goiri, Jordi Guitart, Md. Haque, William Katsak, Kien Le, Thu D. Nguyen, Jordi Torres

slide-1
SLIDE 1

Leveraging Renewable Energy in Data Centers: Present and Future

Ricardo Bianchini

Department of Computer Science

Collaborators: Josep L. Berral, Inigo Goiri, Jordi Guitart, Md. Haque, William Katsak, Kien Le, Thu D. Nguyen, Jordi Torres

slide-2
SLIDE 2

Motivation

  • Data centers = machine rooms to giant warehouses
  • Consume massive amounts of energy (electricity)

30 60 90

2000 2005 2010 Billion KWh/year Electricity consumption of US DCs [JK’11]

30 60 90 120 150 180 210 240 270

2000 2005 2010 Billion KWh/year Electricity consumption of WW DCs [JK’11] 2% 1.5%

slide-3
SLIDE 3

Motivation

  • Electricity comes mostly from burning fossil fuels

CO2 of world-wide DCs [Mankoff’08] Electricity sources in US & WW [DOE’10]

100 104 108 112 116 120

Nigeria Data Centers Czech Rep.

35th 34th MMT/year

0% 20% 40% 60% 80% 100%

US World

Others Renewables Nuclear Natural Gas Coal

Can we use renewables to reduce this footprint?

slide-4
SLIDE 4

Outline

  • DC energy usage and carbon footprint
  • Reducing footprint with renewables: 2 approaches
  • Our target and research challenges
  • Software for leveraging solar energy
  • Parasol: our solar micro-data center
  • Current and future works
  • Conclusions
slide-5
SLIDE 5

Greening DCs

  • 1. Power purchase agreement, off-site generation

– Renewable energy produced at the best location – Energy losses: ~15% [IEC’07] – Example: Google buys wind power from NextEra

slide-6
SLIDE 6

Greening DCs

  • 1. Power purchase agreement, off-site generation

– Renewable energy produced at the best location – Energy losses: ~15% [IEC’07] – Example: Google buys wind power from NextEra

  • 2. Co-location, self-generation

– Lower peak power, energy costs with self-generation – Location may not be ideal for DC or renewable plant – Examples: MSFT placed DC near a hydro plant in OR Apple built a 40MW solar array in NC

  • No approach is perfect
slide-7
SLIDE 7

Outline

  • DC energy usage and carbon footprint
  • Reducing footprint with renewables
  • Our target and research challenge
  • Software and hardware for leveraging solar energy
  • Current and future works
  • Conclusions
slide-8
SLIDE 8

Our research target

  • Co-location or self-generation with solar and/or wind

– Pros: Clean and available – Cons: Space and cost

slide-9
SLIDE 9

Solar and wind are clean

100 200 300 400 500 600 700 800 900 1000 g CO2e per KWh over lifetime [Sovacool’08]

slide-10
SLIDE 10

100 200 300 400 500 600 700 800 900 1000

Solar and wind are clean

g CO2e per KWh over lifetime [Sovacool’08]

slide-11
SLIDE 11

Solar is more available in the US

Wind Solar [NREL’12]

Fair Good Excellent Outstanding Superb

slide-12
SLIDE 12

Space: Solar PV efficiencies are increasing

[IEA’10] Efficiency rates of PV modules

slide-13
SLIDE 13

5 10 15 20 25

Space: Solar PV capacity factors today

[PVOutput’12]

slide-14
SLIDE 14

Cost of solar PV energy is decreasing

Grid electricity prices have been increasing: 30%+ since 1998 [EIA’12]

[DOE’11,Solarbuzz’12] 4 8 12 16 20 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 2023 2025 2027 2029 2011 Dollars per Watt

Inverters Panels Installed

slide-15
SLIDE 15

4 8 12 16 20 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 2023 2025 2027 2029 [DOE’11,Solarbuzz’12] 2011 Dollars per Watt

Inverters Panels Installed

Cost of solar PV energy is decreasing

spike in demand world-wide recession back to historical levels

slide-16
SLIDE 16

4 8 12 16 20 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 2023 2025 2027 2029 [DOE’11,Solarbuzz’12] 2011 Dollars per Watt

Inverters Panels Installed

Cost of solar PV energy is decreasing

< 1/2 of current cost

With incentives, the installed price can go down by another 40-60%

slide-17
SLIDE 17

Solar space and cost: Present and future

Space as a factor of rack area Present Future (2020-2030) Density per rack 8kW (200W 1U servers) ~47x ~24x 2kW (25W 0.5U servers) ~12x ~6x

Assuming 30% server utilization, 50% solar energy, NJ capacity factor, and 1 row of panels

slide-18
SLIDE 18

Solar space and cost: Present and future

Space as a factor of rack area Present Future (2020-2030) Density per rack 8kW (200W 1U servers) ~47x ~24x 2kW (25W 0.5U servers) ~12x ~6x Cost per AC Watt Present Future (2020-2030) ~$2.30 < $1.20 Time to amortize cost Present Future (2020-2030) ~12 years < 6 years

Assuming 30% server utilization, 50% solar energy, NJ capacity factor, and 1 row of panels Assuming above costs, NJ capacity factor, and NJ grid energy prices Assuming self-generation and federal + state incentives

slide-19
SLIDE 19

Solar space and cost: Present and future

Space as a factor of rack area Present Future (2020-2030) Density per rack 8kW (200W 1U servers) ~47x ~24x 2kW (25W 0.5U servers) ~12x ~6x

Assuming 30% server utilization, 50% solar energy, NJ capacity factor, and 1 row of panels

Wind takes ~12x less space and is ~3x cheaper

Cost per AC Watt Present Future (2020-2030) ~$2.30 < $1.20 Time to amortize cost Present Future (2020-2030) ~12 years < 6 years

Assuming above costs, NJ capacity factor, and NJ grid energy prices Assuming self-generation and federal + state incentives

slide-20
SLIDE 20

Main challenge: Supply of power is variable!

  • Batteries and net metering are not ideal
  • We need to match the energy demand to the supply

Solar power Workload

Now Power

slide-21
SLIDE 21

Main challenge: Supply of power is variable!

  • Many research questions:

– What kinds of DC workloads are amenable? – What kinds of techniques can we apply? – How well can we predict solar energy availability? – If batteries are available, how should we manage them? – Can we leverage geographical distribution?

  • Building hardware & software to answer questions
slide-22
SLIDE 22

Outline

  • DC energy usage and carbon footprint
  • Reducing footprint with renewables
  • Our target and research challenges
  • Hardware and software for leveraging solar energy
  • Current and future works
  • Conclusions
slide-23
SLIDE 23

Green DC software

  • Follow the renewables [HotPower’09, SIGMETRICS’11]
  • Duty cycle modulation with sleep states [ASPLOS’11]
  • Quality degradation for interactive loads [UCB-TR’12]
  • Adapt the amount of batch processing [HotPower’11]
  • Delay batch jobs while respecting deadlines

– GreenSlot [SC’11], GreenHadoop [Eurosys’12]

slide-24
SLIDE 24

Overall “delay-until-green” approach

  • Predict green energy availability

– Weather forecasts

  • Schedule jobs

– Maximize green energy use – If green not available, consume cheap brown electricity

  • May delay jobs but must meet deadlines
  • Send idle servers to sleep to save energy
  • Manage data availability if necessary
slide-25
SLIDE 25

GreenHadoop scheduling

Job3 Job1 Job4 Job5 Job6 Job2

Estimate the energy required by jobs

Job3 Job1 Job4 Job5 Job6 Job2

slide-26
SLIDE 26

GreenHadoop scheduling

Job3 Job1 Job4 Job5 Job6 Job2 Power Time Now

Assign green energy first Predict energy availability (weather forecast)

On-peak Off-peak Off-peak

slide-27
SLIDE 27

GreenHadoop scheduling

Job3 Job1 Job4 Job5 Job6 Job2 Time Now

Assign cheap brown energy

Power Previous peak On-peak Off-peak Off-peak

slide-28
SLIDE 28

GreenHadoop scheduling

Job3 Job1 Job4 Job5 Job6 Job2 Time Now

Assign expensive energy

Power Active servers On-peak Off-peak Off-peak Current power → Active servers

slide-29
SLIDE 29

GreenHadoop scheduling

Time Now Active servers

As time goes by… the number of active servers changes

Power

slide-30
SLIDE 30

Brown consumed Green consumed Green produced Brown price

31% more green 39% cost savings

GreenHadoop for Facebook workload

Brown consumed Green consumed Green predicted Brown price Hadoop GHadoop

slide-31
SLIDE 31

Outline

  • DC energy usage and carbon footprint
  • Reducing footprint with renewables
  • Our target and research challenges
  • Software and hardware for leveraging solar energy
  • Current and future works
  • Conclusions
slide-32
SLIDE 32

The Rutgers Parasol Project

slide-33
SLIDE 33

Parasol: Our hardware prototype

  • Unique research platform

– Solar-powered computing – Remote DC deployments – Software to exploit renewables within and across DCs – Tradeoff between renewables, batteries, and grid energy – Free cooling, wimpy servers, solid-state drives

slide-34
SLIDE 34

Parasol details

  • Steel structure on the roof

– Container hosts 2 racks of IT – 16 solar panels: 3.2 kW peak

  • Backup power

– Batteries and power grid

  • IT equipment

– 64 Atom servers (so far): 1.7 kW

  • Cooling

– Free cooling: 10 -- 400 W – Air conditioning: 2 kW

slide-35
SLIDE 35

Outline

  • DC energy usage and carbon footprint
  • Reducing footprint with renewables
  • Our target and research challenges
  • Software and hardware for leveraging solar energy
  • Current and future works
  • Conclusions
slide-36
SLIDE 36

Current and future works

  • Provisioning the solar array and batteries
  • Free cooling and its costs/benefits, world-wide
  • DC placement with probabilistic green energy guarantees
  • GreenNebula: follow the renewables
  • HotPower’09, IGCC’10, SC’11, EuroSys’12, IGCC’13, ASPLOS’13
slide-37
SLIDE 37

Conclusions

  • Reduce the carbon footprint of ICT, data centers
  • Topic is interesting and has societal impact
  • Prior work on software and hardware
  • Lots left to do…

http://parasol.cs.rutgers.edu