1 of 20 Low Power and Dynamic Optimization Techniques for Power-Constrained Domains Ann Gordon-Ross Department of Electrical and Computer Engineering University.

Slides:



Advertisements
Similar presentations
Reducing Network Energy Consumption via Sleeping and Rate- Adaption Sergiu Nedevschi, Lucian Popa, Gianluca Iannaccone, Sylvia Ratnasamy, David Wetherall.
Advertisements

Zhiguo Ge, Weng-Fai Wong, and Hock-Beng Lim Proceedings of the Design, Automation, and Test in Europe Conference, 2007 (DATE’07) April /4/17.
1 A Self-Tuning Configurable Cache Ann Gordon-Ross and Frank Vahid* Department of Computer Science and Engineering University of California, Riverside.
Improving TCP Performance over Mobile Ad Hoc Networks by Exploiting Cross- Layer Information Awareness Xin Yu Department Of Computer Science New York University,
Helper Protocols Protocols that either make it easier for IP to do its job, or extend the capabilities of the network layer.
UNIT-IV Computer Network Network Layer. Network Layer Prepared by - ROHIT KOSHTA In the seven-layer OSI model of computer networking, the network layer.
1 Cross-Layer Scheduling for Power Efficiency in Wireless Sensor Networks Mihail L. Sichitiu Department of Electrical and Computer Engineering North Carolina.
1 Soft Timers: Efficient Microsecond Software Timer Support For Network Processing Mohit Aron and Peter Druschel Rice University Presented By Jonathan.
A Self-Tuning Cache Architecture for Embedded Systems Chuanjun Zhang, Vahid F., Lysecky R. Proceedings of Design, Automation and Test in Europe Conference.
Choosing Beacon Periods to Improve Response Times for Wireless HTTP Clients Suman Nath Zachary Anderson Srinivasan Seshan Carnegie Mellon University.
CS335 Networking & Network Administration Tuesday, May 11, 2010.
Introduction to Management Information Systems Chapter 5 Data Communications and Internet Technology HTM 304 Fall 07.
A One-Shot Configurable- Cache Tuner for Improved Energy and Performance Ann Gordon-Ross 1, Pablo Viana 2, Frank Vahid 1, Walid Najjar 1, and Edna Barros.
Automatic Tuning of Two-Level Caches to Embedded Applications Ann Gordon-Ross and Frank Vahid* Department of Computer Science and Engineering University.
1 Hot Caches, Cool Techniques: Online Tuning of Highly Configurable Caches for Reduced Energy Consumption Ann Gordon-Ross Department of Computer Science.
Exploring the Tradeoffs of Configurability and Heterogeneity in Multicore Embedded Systems + Also Affiliated with NSF Center for High- Performance Reconfigurable.
11 NETWORK PROTOCOLS AND SERVICES Chapter 10. Chapter 10: Network Protocols and Services2 NETWORK PROTOCOLS AND SERVICES  Identify how computers on TCP/IP.
CCNA Guide to Cisco Networking Fundamentals Fourth Edition
Guide to MCSE , Second Edition, Enhanced1 Windows XP Network Overview Most versatile Windows operating system Supports local area network (LAN) connections.
1 Internetworking : Internet architecture and TCP/IP Protocol TK3133 Computer Networking Technology.
CPACT – The Conditional Parameter Adjustment Cache Tuner for Dual-Core Architectures + Also Affiliated with NSF Center for High- Performance Reconfigurable.
Tufts Wireless Laboratory School Of Engineering Tufts University “Network QoS Management in Cyber-Physical Systems” Nicole Ng 9/16/20151 by Feng Xia, Longhua.
TCP/IP. The Internet Protocol Suite is the set of communications protocols used for the Internet and other similar networks. It is commonly also known.
LWIP TCP/IP Stack 김백규.
TCP/IP: Basics1 User Datagram Protocol (UDP) Another protocol at transport layer is UDP. It is Connectionless protocol i.e. no need to establish & terminate.
Low-Power Wireless Sensor Networks
Examining TCP/IP.
Energy Efficient Digital Networks Rich Brown Lawrence Berkeley National Laboratory Presentation to DOE State Energy Advisory Board Meeting August 14, 2007.
IP Forwarding.
Chap 9 TCP/IP Andres, Wen-Yuan Liao Department of Computer Science and Engineering De Lin Institute of Technology
1 © 2003, Cisco Systems, Inc. All rights reserved. CCNA 2 Module 9 Basic Router Troubleshooting.
Vladimír Smotlacha CESNET Full Packet Monitoring Sensors: Hardware and Software Challenges.
1 of 20 Phase-based Cache Reconfiguration for a Highly-Configurable Two-Level Cache Hierarchy This work was supported by the U.S. National Science Foundation.
CCNA 3 Week 4 Switching Concepts. Copyright © 2005 University of Bolton Introduction Lan design has moved away from using shared media, hubs and repeaters.
CHAPTER 5 TCP/IP PROTOCOLS. P ROTOCOL STANDARDS Protocols are formal rules of behavior When computers communicate, it is necessary to define a set of.
A Single-Pass Cache Simulation Methodology for Two-level Unified Caches + Also affiliated with NSF Center for High-Performance Reconfigurable Computing.
Lecture (Mar 23, 2000) H/W Assignment 3 posted on Web –Due Tuesday March 28, 2000 Review of Data packets LANS WANS.
TCP/IP (Transmission Control Protocol / Internet Protocol)
BZUPAGES.COM Presentation on TCP/IP Presented to: Sir Taimoor Presented by: Jamila BB Roll no Nudrat Rehman Roll no
Dynamic Phase-based Tuning for Embedded Systems Using Phase Distance Mapping + Also Affiliated with NSF Center for High- Performance Reconfigurable Computing.
A Resource Efficient Content Inspection System for Next Generation Smart NICs Karthikeyan Sabhanatarajan, Ann Gordon-Ross* The Energy Efficient Internet.
Real-Time Performance Analysis of Adaptive Link Rate Baoke Zhang, Karthikeyan Sabhanatarajan, Ann Gordon-Ross*, Alan D. George* This work was supported.
Analysis of Cache Tuner Architectural Layouts for Multicore Embedded Systems + Also Affiliated with NSF Center for High- Performance Reconfigurable Computing.
Thermal-aware Phase-based Tuning of Embedded Systems + Also Affiliated with NSF Center for High- Performance Reconfigurable Computing This work was supported.
1 of 20 Smart-NICs: Power Proxying for Reduced Power Consumption in Network Edge Devices Karthikeyan Sabhanatarajan, Ann Gordon-Ross +, Mark Oden, Mukund.
Linux Operations and Administration Chapter Eight Network Communications.
GreenCloud: A Packet-level Simulator of Energy-aware Cloud Computing Data Centers Dzmitry Kliazovich ERCIM Fellow University of Luxembourg Apr 16, 2010.
Lightweight Runtime Control Flow Analysis for Adaptive Loop Caching + Also Affiliated with NSF Center for High- Performance Reconfigurable Computing Marisha.
Exploiting Dynamic Phase Distance Mapping for Phase-based Tuning of Embedded Systems + Also Affiliated with NSF Center for High- Performance Reconfigurable.
TCP/IP1 Address Resolution Protocol Internet uses IP address to recognize a computer. But IP address needs to be translated to physical address (NIC).
The OSI Model. Understanding the OSI Model In early 1980s, manufacturers began to standardize networking so that networks from different manufacturers.
Address Resolution Protocol Yasir Jan 20 th March 2008 Future Internet.
IP - Internet Protocol No. 1  Seattle Pacific University IP: The Internet Protocol Kevin Bolding Electrical Engineering Seattle Pacific University.
Ad Hoc – Wireless connection between two devices Backbone – The hardware used in networking Bandwidth – The speed at which the network is capable of sending.
CompTIA Security+ Study Guide (SY0-401)
Topics discussed in this section:
ARP and RARP Objectives Chapter 7 Upon completion you will be able to:
Net 323: NETWORK Protocols
CompTIA Security+ Study Guide (SY0-401)
Anne Pratoomtong ECE734, Spring2002
Ultra-Low Duty Cycle MAC with Scheduled Channel Polling
Tosiron Adegbija and Ann Gordon-Ross+
Ann Gordon-Ross and Frank Vahid*
Tosiron Adegbija and Ann Gordon-Ross+
A Self-Tuning Configurable Cache
Realizing Closed-loop, Online Tuning and Control for Configurable-Cache Embedded Systems: Progress and Challenges Islam S. Badreldin*, Ann Gordon-Ross*,
Automatic Tuning of Two-Level Caches to Embedded Applications
Dynamic Power Management for Streaming Data
Presentation transcript:

1 of 20 Low Power and Dynamic Optimization Techniques for Power-Constrained Domains Ann Gordon-Ross Department of Electrical and Computer Engineering University of Florida, Gainesville

2 of 23 2 Power-Constrained Domains Embedded Systems Domain –Increased battery life –Decreased cooling requirements

3 of 23 3 Dynamic Optimizations ?

4 of 23 4 Dynamic Optimizations Benefits No user effort –User doesn’t know what dynamic optimizations are No application-designer effort –Reduced time to market –Reduced NRE costs System reacts to changing environment –Changes in input stimulus –Changes in software/firmware Reduced power consumption –Longer battery life –Reduced cooling requirements

5 of 23 5 Cache Power Consumption Memory access: 50% of embedded processor’s system power –Caches are power hungry –ARM920T (Segars 01) –M*CORE (Lee/Moyer/Arends 99) Thus, caches are a good candidate for optimizations Main Mem L1 Cache Processor L2 Cache 53%

6 of 23 6 Reducing Cache Energy Consumption Different applications have vastly different cache requirements –Total size, line size, and associativity Cache parameters that don’t match an application’s behavior can waste over 60% of energy (Gordon-Ross 05) 4KB 16 byte 2-way 2KB 32 byte direct-mapped 8KB 64 byte 4-way

7 of 23 7 Dynamic Cache Tuning Cache tuning is the process of determining the appropriate cache parameters for an application –Requires a tunable cache Cache parameter values can be varied during runtime –Requires tuning hardware Orchestrates cache tuning Energy Executing in base configuration Tunable cache Tuning hw TC Cache Tuning TC Download application Microprocessor Cache energy savings of 62% on average!

8 of 23 8 Dynamic Cache Tuning Reconfigure the cache dynamically to adapt to different phases of program execution or different applications in a multi- application environment Base cache energy Time Energy Consumption Phase-tuned Change cache

9 of 23 9 Dynamic Cache Tuning Challenges Base cache energy Time Energy Consumption Phase Interval Base cache energy Time Energy Consumption Runtime energy Tuning interval Excess tuning energy Tuning interval too short Tuning interval too long Base cache energy Time Energy Consumption Runtime energy Tuning interval Wasted energy in suboptimal configuration Need a good tuning interval –Tuning interval is the time between invocations of the tuning hardware –Should closely match phase interval - length of time the system executes between phase changes Problem: How does the tuning hardware determine when to invoke cache tuning - must have knowledge of the future to obtain optimal results

10 of Periodic System - Fixed Phase Interval Phase interval fixed at 10 million cycles Tuning interval too short Tuning interval too long Energy savings = 32% (includes 7% overhead due to tuning) Base Line Negative savings if tuning interval is greater than phase interval!

11 of Dynamic Cache Tuner Energy Savings Base line Observed similar results for less periodic systems, but still much work to be done. 29% energy savings - within 8% of optimal Normalized Energy

12 of Future Directions Dynamic optimizations in a multi-core environment –Cache hierarchy – some levels may be shared –Dynamic load distribution –Dynamic per-core shutdown or voltage reduction for reduced power consumption –Etc – Many single-core optimizations can be non- trivially applied to a multi-core environment –Dynamic tuning enables energy savings with no extra designer effort – suitable for standard binary situations, changing environment situations, etc.

13 of Power-Constrained Domains INTERNET

14 of Internet Power Consumption Connected edge devices account for 2% of the total power consumed in the US [EPA-06] –130 TWh/Year This is $1.3 $.10 per kWh 1 single-unit nuclear power plant outputs 8 TWh/Year Translates to 16 single-unit nuclear power plants! Why so much power? –PCs can consume up to 200 W –1 billion PCs worldwide by 2010 [Kanellos-04] What can we do? –PCs are idle 75% of the time [Purushothaman-06] –But only 10% of PCs are allowed to sleep during that time [EPA-06] –Sleeping reduces power consumption by 80% or more –If PCs were allowed to sleep, only 3 single-unit nuclear power plants would be required Question: Why aren’t these PCs asleep?!?!

15 of Maintaining Network Connectivity INTERNET IDLE GNUTELLA FILE SHARING APPLICATION FILE QUERY PACKET FILE RESPONSE PACKET Bob Alice Alice checks to see if Bob has a file needed for p2p file sharing Z Z z z FILE QUERY PACKET Problem: PC must be awake to maintain network connectivity

16 of A Solution – Power Proxying Primary challenge is to maintain network connectivity while the PC is power down to standby mode - sleeping Some packets do not require a complex response –Automated responses are sufficient –Network Interface Card (NIC) can act as proxy for the PC –Allow the PC to sleep while NIC services packets with automated responses –A technique known as power proxying –We call such a NIC a “Smart”-NIC - SNIC

17 of Power Proxying INTERNET IDLE GNUTELLA FILE SHARING APPLICATION Alice Bob Z Z z z PC delegates power to the SNIC to handle to network traffic FILE QUERY PACKET FILE RESPONSE PACKET

18 of Power Proxying INTERNET IDLE Proxiable Packet Response Z Z z z Chatter Packet Non-Proxiable/Wake up Packet SNIC Response Bob

19 of What to Proxy? - Proxiable Protocols Proxiable protocols - Network protocols amenable to proxying –Responses may be automated –Keep alive packets, IP conflict avoidance, etc. Z Z z z IDLE FOUR Categories of Proxiable Packets ARP QUERY ARP RESPONSE PING PING RESPONSE P2P FILE QUERY P2P RESPONSE Mail Notification ARP (Address Resolution Protocol) ICMP (Internet Control Message Protocol) TCP (Transmission Control Protocol) UDP (User Datagram Protocol) What application support is needed to increase sleep time?

20 of Network Slowdown Link rates are increasing to meet network traffic demands –10 Gbps soon to be common place Power of these links increase exponentially However, research shows that links are largely underutilized –1-5% for 1 Gbps –Need high speed for traffic bursts During times of light utilization, don’t need full link speed –i.e. Adaptive Link Rate –Change link speed to meet traffic demands –i.e. Switch between 100 Mbps and 1 Gbps –IEEE 802.3az task force established in Nov 2006

21 of Network Slowdown Which components within the NIC (or edge device i.e. first level switches and routers) can exploit times of low utilization –Reconfigurable switch fabrics –Clock down processor –Different processor sleep levels –Disable links –Reconfigurable buffers –Etc

22 of Network Slowdown Challenges –Network devices do not contain hardware primitives for low power operation –Designed to operate for peak traffic load Devices are too power hungry during average traffic load –Propose to design network devices optimized for average traffic load, but can handle peak traffic load –What hardware primitives are necessary? Low hardware sleep levels take longer to wake up than link speeds Bursts are unpredictable How can we architect support for these bursts to allow network devices time to wake up?

23 of Conclusions Dynamic optimizations of embedded systems –Self-tuning system for reduced energy and/or power consumption –Determining when to reconfigure is challenging –Move dynamic optimization research into multi-core Internet power consumption –Power proxying Allows host to sleep and maintain network connectivity –Network slowdown Designing network devices optimized for average traffic load