Energy Consumption in Energy Consumption in IP Networks IP - PowerPoint PPT Presentation

Energy Consumption in Energy Consumption in IP Networks IP Networks Rodney S. Tucker, Jayant Baliga, Robert Ayre, Kerry Hinton, Wayne V. Sorin ARC Special Research Centre for Ultra-Broadband Information Networks (CUBIN) University of Melbourne r.tucker@ee.unimelb.edu.au

Energy Consumption of the Network Energy Consumption of the Network Hot spot Why should we be interested in energy? Power In • OPEX • Greenhouse Impact • Managing “Hot Spots” - Getting the energy in - Getting the heat out • Energy-limited capacity bottlenecks • Enabling energy efficiencies in other sectors

Where are We Heading ? Where are We Heading ? � More users � More data-intensive applications, e.g. video � More often and for longer periods Increasing demand → operators provide faster access � and increased core capacity � New applications enabled by faster access Energy Consumption Grows Energy Consumption Grows

Summary Summary � Modeling energy consumption of the Internet - Core, metro, and access networks � Energy in network routers � Energy in optical transmission � Will (can) optical switching technologies help to reduce energy consumption?

What is the Carbon Footprint of Telecoms? Global Telecoms Footprint (devices & infrastructure) Mobile Network Mobile handsets Fixed Narrowband 2002 1450+% growth 20X increase 2020 0 100 200 300 400 Footprint (MtCO 2 p.a.) Broadband Fixed Broadband Modems Adapted from “SMART 2020: Enabling the low carbon economy in the information age,” GeSI, 2008 www.gesi.org

Energy Model of Simple IP Network Energy Model of Simple IP Network Core CRS-1 Core ~ 10 kW / rack Core Core Core Core Packet Core over Sonet Fibre Amps Metro WDM 12816 Edge Edge Edge ~ 4 kW OLT - 100W Curb Curb Curb Curb Access OLT - 100W Passive Optical Network ONU ~ 5-10W 0.1 - 1000 Mb/s to the user Baliga et al., 2007

Number of Hops in the Internet Number of Hops in the Internet 2006 Data 0.1 0.08 ] k = 0.06 AT&T: 20 hops Pr H [ 0.04 0.02 0 0 15 10 20 25 5 Number of Hops, k Source: P. Van Mieghem, “Performance Analysis of Computer Systems and Networks” , Cambridge (2006)

Power Consumption of IP Network Power Consumption of IP Network 25 2007 Technology 20 router hops 20 Contention ratio = 25 1.0 % of Electricity Supply Power (W/user) Total 15 Today’s Internet (~ 2.5 Mb/s) Routers 0.5 10 Access (PON) 5 SDH/WDM Links 0 0 150 0 50 100 200 250 Peak Access Rate (Mb/s) Baliga et al., 2007

Ultra- -Broadband Access Broadband Access Ultra 100 5.0 2007 Technology 20 router hops Contention ratio = 25 % of Electricity Supply Power (W/user) Total 50 2.5 Routers Access (PON) SDH/WDM 0 0 0 1000 500 Peak Access Rate (Mb/s) Baliga et al., 2007

Technology Improvements Technology Improvements 3.0 60 % of Electricity Consumption Efficiency Total Power Per User (W) Improvement Rate = 0% p.a 2.0 40 5% p.a 10% p.a 1.0 20 20% p.a 0 0 1 300 100 200 400 Peak Access Rate (Mb/s) Baliga, et al, 2008, unpublished

Energy Consumption in Access Networks Energy Consumption in Access Networks NEC CM7710T Access N/W Cabinet PON Edge Node Splitter Cisco Cabinet NEC NEC VF200F6 12816 CM7700S FTTN with VDSL2 Zyxel VES-1616F-34 Cisco NEC GM100 4503 PtP Axxcelera ExcelMax CPE WiMAX Axxcelera ExcelMax BTS

Power Consumption in Access Networks Power Consumption in Access Networks 40 Oversubscription = 10 25 Power Per User (W) WiMAX FTTN 20 PtP PON 0 100 250 1 10 Peak Access Rate (Mb/s) • Wireless access consumes more energy than optical access • PON FTTH is “greener” than FTTN

Network Energy Consumption per Bit Network Energy Consumption per Bit 10 -3 ~100 μ J/b 20 hops 10 -4 ~1 μ J/b Energy per bit (J) Total 10 -5 10 -6 Routers Access (PON) 10 -7 WDM Links 10 -8 25 2500 2.5 250 Peak Access Rate (Mb/s)

Observations Observations • Optical transport (WDM) consumes relatively little energy < 5% of energy > 25% of CAPEX • Access network dominates at low rates – Standby/Sleep mode needed • Network routers dominate at higher rates – Need to reduce hop count • improve router efficiency • manage distribution and replication of content (IPTV) •

Power Consumption in Routers Power Consumption in Routers 1,000,000 ? P = C 2/3 100,000 where P is in Watts Power consumption (W) where C is in Mb/s 10,000 10 nJ/bit 1,000 P ~ 10 100 100 nJ/bit 10 1 1 Mb/s 1 Tb/s 1 Pb/s 1 Gb/s Router Throughput Source: METI, 2006, Nordman, 2007

Energy Bottleneck Energy Bottleneck High-end router: Cisco CRS-1 Linecard Chassis Switch Fabric Chassis: Fully equipped: Capacity: 0.64 Tb/s Power: 8 kW Multi-rack router Power: 13.6 kW Capacity: 41 Tb/s Power ~ 1 MW X2 every 18 months Per Rack Source: Neilsen, 2006; Deutche Telekom, 2007

Electronic Routers Electronic Routers Forwarding Engine Fibers J Switch Fabric Demutiplexers Multiplexers O/E Buffers Switch Fabrics Converters Reduced bit rate (i.e. parallel processing) Electronics Optics

Energy in Electronic and Optical Routers Energy in Electronic and Optical Routers Line Card Data Plane Buffer Routing Tables Switch Forwarding Engine O/E Power Fabric supply Buffer inefficiency I/O Routing Switch Buffer Engine Control Fans and Forwarding Engine O/E blowers Control Plane Energy/bit Total Electronic 1.1 nJ 1.0 nJ 3.5 nJ 10 nJ 0.7 nJ 3.2 nJ 0.5 nJ (2008) Electronic 20 pJ 25 pJ 80 pJ 50 pJ 250 pJ 65 pJ 10 pJ (2018) Optical 25 pJ 15 pJ 80 pJ 0 65 pJ 15 pJ ? 200 pJ (2018) Optical Packet Switching is not a promising alternative G. Epps, Cisco, 2007, ITRS, 2005, R. Tucker, JLT, 2006

Contention Resolution in the Wavelength Domain Contention Resolution in the Wavelength Domain λ Forwarding Engine 1 Switch Fabric Forwarding Engine λ Forwarding Engine n Fatal Flaw: Require large n for low blocking probability ( n ~3 -10 x) Wong, JLT 2006 Pathiban et al., JLT 2009 100 Total (Wavelength-domain contention resolution ) Power (W/user) Routers 1.2 X Total (Conventional router) Routers 50 WDM 5 X Access WDM 0 0 500 1000 Peak Access Rate (Mb/s)

The Challenge The Challenge 10 % - 20 % p.a. continuous improvement in efficiency 3.0 60 % of Electricity Consumption Efficiency Total Power Per User (W) Improvement Rate = 0% p.a 2.0 40 5% p.a 10% p.a 1.0 20 Target 20% p.a 0 0 1 300 100 200 400 Peak Access Rate (Mb/s) Baliga, et al, 2008, unpublished

Summary Summary • Energy consumption currently dominated by the access network • The energy bottleneck in routers is looming - More significant than the so-called “electronic speed bottleneck” • Key strategies for efficient network design - Control energy in the access network (e.g. sleep mode in modems) - Reduce the hop count (i.e. “agile” optical bypass) - Caching and content distribution networks - Continuous improvement in router efficiency