Performance of Benchmarks on the Cell Processor

Size: px
Start display at page:

Download "Performance of Benchmarks on the Cell Processor"

Transcription

1 Performance of Benchmarks on the Cell Processor Kerry Creeron Michael Smith Dan Perron CS/ECE 752 Advanced Computer Architecture I The University of Wisconsin-Madison ktcreeron at wisc dot edu, masmith27 at wisc dot edu, djperron at wisc dot edu Abstract The Cell processor represents a $4 million dollar investment on the part of Sony, Toshiba, and IBM, which is poised to make its way into a slew of embedded devices, from High Definition TV s, to cellular phones. Given that the Cell is relatively new, and its performance is currently unknown, we propose to better gauge its capabilities by modifying portions of the MiBench suite for use on the Cell. In the end, we find that performance speedup using the cell is possible, but that the processor s architectural limitations and thread creation overhead make reaping such benefits difficult. 1. Introduction New processor architectures are nothing unusual in this day in age. The past few years have seen a large number of new architectures, including Sun s Niagara and Intel s Core architectures. And with the introduction of these new architectures, one sees rather obviously, the trend that processors are becoming increasingly parallel in nature. The name of the game seems to more cores. The Cell processor, developed by Sony, Toshiba, and IBM (STI), is a new take on the multicore approach. STI s implementation features a simple, inorder PowerPC core, known as the PowerPC Processing Element (PPE) with 8 Synergistic Processing Elements (SPE s) designed to perform computationally intensive tasks, such as floating point operations and vector operations. The PPE is connected to the 8 SPE s via a ring interconnect, known as the Element Interconnect Bus (EIB). Transfers to and from the PPE and SPE s are performed using DMA, the benefit being that no computation is wasted coordinating memory transfers among the various elements of the chip. The twist in the Cell architecture lies in the fact that while the memory is visible to all the elements on the chip, memory transfers must be explicitly scheduled between the various elements. Thus, memory must be carefully managed to ensure coherence is maintained. Given the departure of the Cell from a truly shared memory architecture, its rather explicit take on parallelizing code, and fact these three computing giants intend to use the Cell in a wide variety of applications, from the Playstation 3, to High Definition TV s, to cellular phones, we believe that it is important to gauge the suitability of the Cell in such embedded applications so that systems designers can make more informed decisions, as opposed to the current cloud of uncertainty hanging over the Cell s performance. To benchmark the Cell, we propose to use the freely available, open source, MiBench suite, a suite designed for embedded applications to gauge the performance and scalability of the Cell. Of this suite, we selected the automotive area as a prime candidate for running on the

2 Cell, due to the parallel nature of the code found in these individual programs, and the lack of data dependencies. In order to more accurately characterize the performance of the Cell s unique architecture, we extensively modify the MiBench code so that computation is farmed out to each of the SPE s. In addition, we are aware of the fact that yield is currently an issue with the Cell, and as a result, the Playstation 3 s Cell chips come with one of the SPE s disabled. To measure the performance impact of having reduced numbers of SPE s, we vary the number of SPE s the benchmarks are allowed to use by limiting SPE thread creation. Finally, we compare the performance of a Cell Blade server versus an Intel x86 system, and use IBM s systemsim simulator to collect more detailed performance statistics about the SPE s. After modifying the automotive section of MiBench and running our test programs on our test systems, we found that in some instances, typically those where the SPE s had large data sets and each SPE spent a large amount of time performing computationally expensive tasks, the Cell Blade outperformed our baseline x86 system. When data sets were limited and the SPE s performed a smaller number of iterations however, using a larger number of SPE s resulted in decreased performance as compared to using a number of SPE s. This degraded performance is due primarily to the overhead associated with creating a new thread on each SPE, and the DMA transfers involved therein. The remainder of this paper is organized as follows: Section 2 describes our benchmark design. Section 3 describes our benchmark implementation and methodology. Section 4 discusses our performance results. Section 5 discusses related work, and Section 6 concludes the paper. 2. Design To test the cell s performance while running an embedded systems workload we needed in appropriate benchmarking suite to use as a template for our modified versions designed to run on the cell. The performance metrics we chose to look at were both wall clock time and CPI. Wall clock time was chosen in order to get a measurement of cross architecture performance. We also wanted to see how big of a difference there was in changing the number of SPE s used at a time. We chose to do this by analyzing the CPI of the SPE s Benchmark Suite Initially when searching for benchmarking suites we came across EEMBC which is the standard for embedded benchmarking programs. The problem with EEMBC is that there are licensing issues involved in using it and we did not have the proper access. We then came across an open source alternative known as MiBench from the University of Michigan. MiBench consists of several different categories of benchmarks. Those categories include automotive, consumer, network, office, security, and telecomm. The automotive suite was chosen for its small code size, and ease of parallelization. Inside the automotive suite there were 4 programs. These four programs were titled basicmath, bitcount, qsort, and susan. We chose to implement the first three since we were unsure on how best to parallelize susan which is an image recognition

3 program. Basicmath consisted of four functions, a cubic solver, square root solver using bit operations, a degree to radians function, and radians to degrees function, each of these functions were called in a series of for loops with varying pieces of data. Bitcount implemented 7 different bit counting algorithms, each were run on large sets of data. Qsort ran c s qsort function on large amounts of data Performance Analysis We wanted to compare performance both against other architectures as well as on the number of SPE s used. In order to judge performance across architectures we used wall clock time of the entire program using. We tested the code on an x86 based machine as well as on a cell blade server which we had limited access to. In order to test the performance with varying SPE s the IBM cell simulator was used. The metric we chose to compare was CPI. The performance monitor was started right before the program was run and observed immediately after. 3. Implementation In order run the MiBench benchmarks on the Cell processor several changes had to be made to the code. The MiBench suite was eventually treated more like a guideline for writing embedded benchmarks then anything else. Various versions of the benchmarks were created in order to analyze performance on the x86, as well as 1 SPE, 2 SPE, 4 SPE, and 8 SPE versions. 3.1 Benchmark Modifications In order to efficiently parallelize these benchmarks some modifications had to be made. In basicmath the number of loop iterations was changed to be a multiple of 8. This allowed for us to easily split up the program to run on 8, 4, 2 or 1 SPE. From here code was added in to be able to create SPE threads instead of function calls. Each for loop was split up into a certain number of partitions depending on the number of SPE s used. These partitions then sent both a starting number and a number of iterations to the SPE or the entire set of data for that particular partition to the SPE. Bitcount was split up in a similar way. Each bit counting algorithm was ran for a large number of iterations. These iterations were split up amongst the number of SPE s and the data and number of iterations was sent to each SPE. Qsort was parallelized by sending a partition of the data to each SPE to be quick sorted. These sorted partitions were then sent back to the PPE where they were merge sorted. 3.2 Performance Analysis Methods In order to judge performance based on wall clock time two systems were used. Wall clock time was found by calling get time of day at the beginning of the code running in the PPE in order to get an overall execution time. Two different cpu architectures were analyzed. The cpu specifications can be found in table 3.1 Table 3.1 System Specifications System 1 System 2 CPU Pentium 4 Two Cell Processors Frequency 3.ghz 3.2 GHz Ram 1gb 512mb per processor

4 The wall clock time for the cell programs also changed the number of SPE s used. The x86 version of our code did not use threads but instead just normal function calls. In order to compare performance of just changing the number of SPE s the IBM cell simulator was used. The simulator allowed us to get detailed statistics about individual SPE s we chose to look at CPI specifically as well as looking at where the stalls are in the SPE execution. The CPI was only found for 8 and 1 SPE s total, also we had to use modified versions of the benchmarks we ran to get wall clock time. Specifically we had to modify them to only create one thread in the case of 1 SPE or 8 when using the 8 SPE version. This is because the simulator did not have an option to turn off certain SPE s and chose to send threads to different SPE s each time. This made getting CPI on anything less then 8 SPE s almost impossible. 4. Results The first series of results that we present are comparisons of wall clock time of the Cell Blade server, using differing numbers of SPE s, to the wall clock time of the Intel Pentium 4 baseline system. Benchmarks are for the Bit Count Program, Basic Math program, and the Quick Sort program. Time (seconds) Time (Microseconds) Bit Count Bit Count Times 1 SPE 2 SPE 4 SPE 8 SPE x86 Figure 1.1 Basic Math Basicmath Times 1 SPE 2 SPE 8 SPE x86 Figure 1.2 Quick Sort Wall Time of qsort on Pentium 4 vs Cell Blade Time (usecc) Cell Blade Pentium 4 Figure 1.3 Processor Our second series of data compares the number of cycles that each SPE

5 performs, scaling the number of active SPE s from 1 to 8. This data was obtained using IBM s systemsim simulator. In this series, the benchmarks run are decomposed to some extent. While the bit count program is run in its entirety, the basic math is broken down into two of its constituent programs, usqrt, which performs an inverse binary square root, and deg2rad, which performs a radian to degree conversion, to provide a more detailed look of the program, and also to allow for the simulation to complete in a more expedient fashion. Finally, we were unable to obtain reliable numbers from the systemsim simulator for the quicksort benchmark and as such, it was omitted. Number of Cycles Cycles Bit Count Bit Count Cycles 1 SPE 8 SPEs Figure 2.1 Usqrt USqrt 1 SPE 8 SPE Figure 2.2 Cycles Deg2rad Deg2Rad 1 SPE 8 SPE Figure 2.3 The first series of data is very telling. The first figure, 1.1, shows that significant performance gains are possible over our baseline Pentium 4 system for the bit count program when 4 or 8 SPE s are used. This is an encouraging sign, showing both that the Cell can scale, and also that it can outperform a more general purpose CPU in some situations. Figures 1.2 and 1.3 paint a very different picture. Here, we not only see the number of SPE s hurting performance, but we also see the Cell losing to a Pentium 4 system with a lower clock speed. The increase in wall clock time can be easily attributed to the overhead associated with thread creation and the additional data transfers that must take place on the EIB in order to move the data to a larger number of SPE s. Additionally, the huge discrepancy between the Cell s performance and the Pentium 4 s can be attributed to the fact that our naïve Pentium 4 implementation was single threaded, so many of the functions could be inlined and no thread overhead was introduced, giving the Pentium 4 a huge boost in performance. Moving on to the second data series, figures 2.1 and 2.2 illustrate the kinds of gains that can be reaped by using multiple SPE s. Clearly the bit count

6 program is amenable to parallelization, as it evinces a 9 times decrease in the number of cycles run on each SPE. The same goes for Usqrt, in which a 5 times decrease in clock occurs between the 1 SPE version and the 8 SPE version. Figure 2.3 illustrates that using more SPE s can incur additional overhead, seeing as the 8 SPE version of the deg2rad program incurs a 2.5 times increase in cycle count. 5. Related Work The key piece of previous work that we relied on for this paper was MiBench: A free, commercially representative embedded benchmark suite. [3] We used this paper and the source code provided with it to create our benchmarks. This paper details the various areas of the MiBench suite, and gives motivation for its suitability as an embedded benchmark suite, as well as for its replacing the EEMBC suite. Additionally, our paper follows in the footsteps of a number of other papers that have worked on testing the Cell s ability to handle various algorithms. The Feldkamp algorithm [1], OpenSSL [2], and Molecular Dynamics [4] have all been successful implementations of algorithms on the Cell. [1] managed to realize a speed gain of more than 2 times using 1 SPE as compared to just the PPE. In [2] a 7.5 times speedup was achieved using 6 SPE s rather than just the PPE. And finally, [4] netted a speedup of 6.1 using 16 SPE s versus 1 SPE. These papers are in line with some of our results, and show that large amounts performance speedup and scalability are achievable using the Cell architecture. CPU. We find this to be necessary because while comparing the cell to itself is useful in determining how well code can be parallelized, it does not indicate with any sort of certainty, the relative merits of the Cell s architecture for a given application. 6. Conclusion At the end of the day, our experience with the Cell processor was something of a mixed bag. While we were able to achieve modest performance increases and demonstrate the scalability of the SPE s, we were unable to do this on a consistent basis, owing to the overhead of thread creation, and lack of computational intensity in some of these programs. Furthermore, the Cell lost in most benchmarks to a Pentium 4 system that costs significantly less. 6.1 Problems We encountered a number of problems throughout the course of our research. First off, the Cell simulator is rather inadequate. It primarily serves as a functional simulator, as opposed to a performance simulator. As such, we based much of our research off of the results output by the simulator. Once we actually moved our code from the simulator onto the production blade server we had access to however, our techniques and strategies for coding on the Cell proved to be inappropriate for the real world. Had we had more time on the Cell blade, we feel confident that we could have tuned our code to realize gains in performance that other research has been able to glean. The main distinction that we make in this paper is to compare the performance of the Cell against another

7 6.2 Future Work Much work is left to be done on the Cell in the embedded arena. There are a number of areas that we would have liked to have touched upon, but did not have the time to do so. Intrinsics, the c wrappers for vector instructions, may provide the performance boost we were hoping to see. After seeing the impact of DMA overhead on our benchmarks, it would also be interesting to see how implementing double buffering would help to reduce memory transfer penalties. Finally, after viewing the overhead associated with thread creation, there is much work to be done in investigating how the thread scheduler on the Cell performs. One option avenue to investigate would be seeing how the SPE s perform when fewer thread joins were performed. Currently, the SPE s run one function, and then a thread_wait is called. What could be done instead is to call thread_wait only at the end of the entire program, testing to see how much amortization can be achieved through more infrequent use of threading commands. 6.3 Final Words The Cell processor presents many unique challenges in the programming space, while at the same time providing great opportunities to see massive performance gains. We believe that our research shows these possibilities, but also provides a realistic characterization of the pitfalls that programmers will face as the Cell continues to mature as an architecture. Even with our ambivalent results, we feel that our insight has been to compare the Cell s performance against dissimilar architectures, and also to demonstrate the scalability and limitations of the architecture, where so many others have taken a blind affection toward this new architecture. In the end, the programming model is a difficult one, the Cell s performance benefits can be difficult to achieve, but given enough time and careful consideration invested, the merits of the Cell can shine through. 6.4 References [1] M. Sakamoto, H. Nishiyama, H. Satoh, S. Shimizu, T. Sanuki, K. Kamijoh, A. Watanabe, A. Asahara, An Implementation of the Feldkamp Algorithm for Medical Imaging On Cell. IBM White Paper. May, 25. [2] N. Costigan, M. Scott, Accelerating SSL using the Vector processors in IBM s Cell Broadband Engine for Sony s Playstation 3TM. Cryptology eprint Archive, Report 27/61. Febuary, 27. [3] M. R. Guthaus, J. S. Ringenberg, D. Ernst, T. M. Austin, T. Mudge, R. B. Brown, MiBench: A free, commercially representative embedded benchmark suite. IEEE 4th Annual Workshop on Workload Characterization. December, 21. [4] G. De Fabritiis, Performance of the Cell processor for biomolecular simulations. Comp. Phys. Commun. (27). March, 27.

The Cell Processor Computing of Tomorrow or Yesterday?

The Cell Processor Computing of Tomorrow or Yesterday? Computing of Tomorrow or Yesterday? Torsten Höfler Department of Computer Science TU Chemnitz 28th February 2006 Outline 1 The Vision 2 3 4 Cell Applications Conclusion The Researchers started in mid 2000

More information

Programming the Cell BE

Programming the Cell BE Programming the Cell BE Max Schneider Chair of Computer Science 3 - Computer Architecture Friedrich-Alexander-University Erlangen-Nuremberg October 26, 2010 Max Schneider (FAU) PParRA October 26, 2010

More information

Cell Software Model. Course Code: L2T1H1-11 Cell Ecosystem Solutions Enablement. Systems and Technology Group

Cell Software Model. Course Code: L2T1H1-11 Cell Ecosystem Solutions Enablement. Systems and Technology Group Cell Software Model Course Code: L2T1H1-11 Cell Ecosystem Solutions Enablement Class Objectives Things you will learn Cell software considerations including Two levels of parallelism: SIMD and parallel

More information

Creating a Passive Solar Water Heating System

Creating a Passive Solar Water Heating System Creating a Passive Solar Water Heating System December 3, 26 Environmental Science 12 Fall 26 Created by: Dan Jacobs Maria K. Chaconas Introduction Challenge: In this course we all acted as research scientists

More information

An Overview of Procedural Fire. How to model procedurally spreading fire

An Overview of Procedural Fire. How to model procedurally spreading fire An Overview of Procedural Fire How to model procedurally spreading fire Table of Contents How to model procedurally spreading fire...1 1. Introduction...3 2. Simulating Fire...4 3. Implementation...5 3.1.

More information

Regional Training. Seminar. » EasyPower Hands-On» Protective Device Coordination» Arc Flash Hazard Analysis. March 12-16, 2018 Austin, TX

Regional Training. Seminar. » EasyPower Hands-On» Protective Device Coordination» Arc Flash Hazard Analysis. March 12-16, 2018 Austin, TX Regional Training» EasyPower Hands-On» Protective Device Coordination» Arc Flash Hazard Analysis Seminar March 12-16, 2018 Austin, TX DoubleTree Austin EasyPower, state-of-the-art Power System Analysis

More information

Advanced Hardware and Software Technologies for Ultra-long FFT s

Advanced Hardware and Software Technologies for Ultra-long FFT s Advanced Hardware and Software Technologies for Ultra-long FFT s Hahn Kim, Jeremy Kepner, M. Michael Vai, Crystal Kahn HPEC 2005 21 September 2005 HPEC 2005-1 This work is sponsored by the Department of

More information

Using BIM model for Fire Emergency Evacuation Plan

Using BIM model for Fire Emergency Evacuation Plan Using BIM model for Fire Emergency Evacuation Plan Adam Stančík 1,, Roman Macháček 2, and Jiří Horák 1 1 Brno University of Technology, Institute of Computer Aided Engineering and Computer Science, 602

More information

Oracle Real Application Clusters (RAC) 12c Release 2 What s Next?

Oracle Real Application Clusters (RAC) 12c Release 2 What s Next? Oracle Real Application Clusters (RAC) 12c Release 2 What s Next? Markus Michalewicz Oracle Redwood Shores, CA, USA Keywords: RAC, Cloud, Private Cloud, Storage, Network, Development, Market Trends, Future

More information

Calcolatori, Internet e il Web

Calcolatori, Internet e il Web Calcolatori, Internet e il Web Refresher on Computer Fundamentals and Networking History of computers Architecture of a computer Data representation within a computer Computer networks and the Internet

More information

A Smart House in the Context of Internet of Things

A Smart House in the Context of Internet of Things A Smart House in the Context of Internet of Things MARIUS RISTEIU *, REMUS DOBRA *, MONICA LEBA *, ANDREEA IONICA ** * Department of Computer and Electrical Engineering ** Department of Management University

More information

Design of Multi-Sensors Cooperation Control System for Intelligent Building

Design of Multi-Sensors Cooperation Control System for Intelligent Building Energy and Power Engineering, 2017, 9, 495-502 http://www.scirp.org/journal/epe ISSN Online: 1947-3818 ISSN Print: 1949-243X Design of Multi-Sensors Cooperation Control System for Intelligent Building

More information

Designing Nanophotonic Interconnection Networks

Designing Nanophotonic Interconnection Networks Designing Nanophotonic Interconnection Networks Christopher Batten Computer Systems Laboratory Cornell University Workshop on the Interaction between Nanophotonic Devices and Systems December 2010 Cornell

More information

A 0.75 Million Point Fourier Transform Chip for Frequency Sparse Signals

A 0.75 Million Point Fourier Transform Chip for Frequency Sparse Signals A 0.75 Million Point Fourier Transform Chip for Frequency Sparse Signals Ezz El Din Hamed Omid Abari, Haitham Hassanieh, Abhinav Agarwal, Dina Katabi, Anantha Chandrakasan, Vladimir Stojanović International

More information

Product Brief for Alarm Analytics V9.2 April 2010

Product Brief for Alarm Analytics V9.2 April 2010 Product Brief for Alarm Analytics V9.2 April 2010 Alarm Analytics V9.2 Product Brief April 2010 Complies with industry standards EEMUA 191 and OSHA PSM 1910.119 for Alarm Analytics, connects to all major

More information

Patent Liability Analysis

Patent Liability Analysis Patent Liability Analysis Introduction The project selected by group two is a parking garage monitoring system. The system will use ultrasonic object detectors to determine if cars are parked in all spaces

More information

The Importance of Pressure Sensors in HVACR Systems

The Importance of Pressure Sensors in HVACR Systems The Importance of Pressure Sensors in HVACR Systems How TE Connectivity s (TE) M3200 Meets New Performance Standards It s Simple, Reliable, Robust By Devin Brock, Manager for Product Knowledge and Training,

More information

Product Brief for Alarm Analytics V9.3 October 2013

Product Brief for Alarm Analytics V9.3 October 2013 Product Brief for Alarm Analytics V9.3 October 2013 Alarm Analytics V9.3 Product Brief October 2013 Complies with industry standards EEMUA 191 and OSHA PSM 1910.119 for Alarm Analytics, connects to all

More information

The Compact Muon Solenoid Experiment. Conference Report. Mailing address: CMS CERN, CH-1211 GENEVA 23, Switzerland

The Compact Muon Solenoid Experiment. Conference Report. Mailing address: CMS CERN, CH-1211 GENEVA 23, Switzerland Available on CMS information server CMS CR -2016/298 The Compact Muon Solenoid Experiment Conference Report Mailing address: CMS CERN, CH-1211 GENEVA 23, Switzerland 28 October 2016 (v2, 25 November 2016)

More information

Validating Vertical-Hole Diagnostic Models with Measured Sandia Data

Validating Vertical-Hole Diagnostic Models with Measured Sandia Data 12 th Annual Sucker Rod Pumping Workshop Renaissance Hotel Oklahoma City, Oklahoma September 27 30, 2016 Validating Vertical-Hole Diagnostic Models with Measured Sandia Data Victoria Pons, Ph.D. Weatherford

More information

Video Analytics Technology for Disaster Management and Security Solutions

Video Analytics Technology for Disaster Management and Security Solutions FEATURED ARTICLES Disaster Management and Security Solutions for a Safe and Secure Way of Life Video Analytics Technology for Disaster Management and Security Solutions The rising demand for safety and

More information

How Computer Simulation Helps to Design Induction Heating Systems

How Computer Simulation Helps to Design Induction Heating Systems Large Pipe Powder Coating Page 1 ASM 2001 How Computer Simulation Helps to Design Induction Heating Systems Dr. Valentin S. Nemkov, Chief Scientist Mr. Robert C. Goldstein, Research Engineer Mr. Robert

More information

Seven Breakthrough Advantages of New Steam Valve Technology

Seven Breakthrough Advantages of New Steam Valve Technology Seven Breakthrough Advantages of New Steam Valve Technology by Anne-Sophie Kedad-Chambareau, Valve Product Marketing Manager, ASCO Gerry Longinetti, Marketing Manager, ASCO A White Paper From ASCO Valve,

More information

Battery Performance Alert

Battery Performance Alert Battery Performance Alert A TOOL FOR IMPROVED PATIENT MANAGEMENT FOR S UNDER BATTERY ADVISORY VERSION 2.0 Abstract BACKGROUND: In October 2016, St. Jude Medical (now Abbott) issued an advisory on a family

More information

Platform Services BACnet Alarm Management

Platform Services BACnet Alarm Management Description: Guide to understanding and setting up BACnet Alarms, visualizing real-time and historical BACnet alarms using AlarmWorX64 Viewer and logging them to SQL database using AlarmWorX64 Logger OS

More information

Green Engineering. E-Design 100 Section 020. Group 5: The Lemons Submitted to Professor Smita Bharti (missing picture)

Green Engineering. E-Design 100 Section 020. Group 5: The Lemons Submitted to Professor Smita Bharti (missing picture) Green Engineering E-Design 100 Section 020 Group 5: The Lemons Submitted to Professor Smita Bharti (missing picture) By: Bo Hui, 19980119bo@gmail.com Adel Rizk, adelrizk1998@gmail.com Dylan Dirkmaat ddirkmaat@yahoo.com

More information

Alarm Analysis with Fuzzy Logic and Multilevel Flow Models

Alarm Analysis with Fuzzy Logic and Multilevel Flow Models Alarm Analysis with Fuzzy Logic and Multilevel Flow Models Fredrik Dahlstrand Department of Information Technology Lund Institute of Technology Box 118, SE-221 00 Lund, Sweden Phone: +46 46 222 95 10 Fax

More information

Samsung SDS BMS Ver.2.0. Technical Specification

Samsung SDS BMS Ver.2.0. Technical Specification Samsung SDS BMS Ver.2.0 Technical Specification Technical Specifications All rights and title, including all intellectual property rights to Samsung SDS BMS" is retained by Samsung SDS Co, Ltd. ("SDS"),

More information

CHAPTER 7 APPLICATION OF 128 POINT FFT PROCESSOR FOR MIMO OFDM SYSTEMS. Table of Contents

CHAPTER 7 APPLICATION OF 128 POINT FFT PROCESSOR FOR MIMO OFDM SYSTEMS. Table of Contents 88 CHAPTER 7 APPLICATION OF 128 POINT FFT PROCESSOR FOR MIMO OFDM SYSTEMS Table of Contents Page No. 7.0 APPLICATION OF 128 POINT FFT PROCESSOR FOR MIMO OFDM SYSTEMS 89 7.1 Introduction 89 7.2 Types of

More information

SAVE: Self-organizing Air VEnt System

SAVE: Self-organizing Air VEnt System SAVE: Self-organizing Air VEnt System 1. Boise State University. 2. Principal Investigator (PI): Gang-Ryung Uh 3. This draft is original and was not submitted to any prior HERC RFPs 4. Executive Summary:

More information

An improved Algorithm of Generating Network Intrusion Detector Li Ma 1, a, Yan Chen 1, b

An improved Algorithm of Generating Network Intrusion Detector Li Ma 1, a, Yan Chen 1, b An improved Algorithm of Generating Network Intrusion Detector Li Ma 1, a, Yan Chen 1, b 1 Department of Information Engineering Guilin University of Aerospace Technology, Guilin China a E-mail: woshiml@sina.com,

More information

DeltaV Analyze. Introduction. Benefits. Continuous automated DeltaV System alarm system performance monitoring

DeltaV Analyze. Introduction. Benefits. Continuous automated DeltaV System alarm system performance monitoring DeltaV Distributed Control System Product Data Sheet December 2016 DeltaV Analyze Continuous automated DeltaV System alarm system performance monitoring Trends of alarms, events, and user actions Identification

More information

The 7 attributes of highly effective quick disconnect couplings for liquid cooling of computers and servers

The 7 attributes of highly effective quick disconnect couplings for liquid cooling of computers and servers WHITE PAPER The 7 attributes of highly effective quick disconnect couplings for liquid cooling of computers and servers By Dennis Downs Business Unit Manager Colder Products Company St. Paul, Minnesota

More information

Centroid Snet 2. Battery Management Software. User Manual V1.1. Eagle Eye Power Solutions, LLC Keeping an Eye on Your Critical Power!

Centroid Snet 2. Battery Management Software. User Manual V1.1. Eagle Eye Power Solutions, LLC Keeping an Eye on Your Critical Power! Eagle Eye Power Solutions, LLC Keeping an Eye on Your Critical Power! Centroid Snet 2 Battery Management Software User Manual V1.1 www.eepowersolutions.com Tel: 1-877-805-3377 info@eepowersolutions.com

More information

Building and Home Improvement Products Distribution Market Report - UK Analysis

Building and Home Improvement Products Distribution Market Report - UK Analysis Building and Home Improvement Products Distribution Market Report - UK 2017-2021 Analysis Published: 27/01/2017 / Number of Pages: 104 / Price: 845.00 Introduction and Overview The 8th edition of the 'Building

More information

System Description AutroMaster V Presentation System

System Description AutroMaster V Presentation System System Description AutroMaster V Presentation System AMASTERV-SYSTEM/XGB, Rev. B, 2018-05-07, Autronica Fire And Security AS COPYRIGHT This publication, or parts thereof, may not be reproduced in any form,

More information

Floating Point Fast Fourier Transform v2.1. User manual

Floating Point Fast Fourier Transform v2.1. User manual User manual Introduction The Fast Fourier Transform (FFT) is an efficient algorithm for computing the Discrete Fourier Transform (DFT). This Intellectual Property (IP) core was designed to offer very fast

More information

DESIGN FOR ELDERLY EGRESS IN FIRE SITUATIONS

DESIGN FOR ELDERLY EGRESS IN FIRE SITUATIONS DESIGN FOR ELDERLY EGRESS IN FIRE SITUATIONS Lauren Folk i, John Gales ii, Steven Gwynne iii and Michael Kinsey iv i Research Student, Architectural Conservation and Sustainability Engineering, Carleton

More information

E. Elnahrawy, X. Li, and R. Martin

E. Elnahrawy, X. Li, and R. Martin Using Area-based Presentations and Metrics for Localization Systems in Wireless LANs E. Elnahrawy, X. Li, and R. Martin Rutgers U. WLAN-Based Localization Localization in indoor environments using 802.11

More information

A Review of a White Paper on Residential Fire Sprinklers

A Review of a White Paper on Residential Fire Sprinklers Advocacy Group A Review of a White Paper on Residential Fire Sprinklers The Advocacy Group has reviewed the white paper entitled Residential Fire Sprinklers For Life Safety by the National Fire Sprinkler

More information

Basic Input Data Needed to Develop a High Performance Fan/Blower with Low Noise, Energy Saving and High Efficiency

Basic Input Data Needed to Develop a High Performance Fan/Blower with Low Noise, Energy Saving and High Efficiency Basic Input Data Needed to Develop a High Performance Fan/Blower with Low Noise, Energy Saving and High Efficiency Haiye Lou TurboMoni Applied Dynamics Lab Abstract It is often happened, when a project

More information

A Comprehensive Approach to Leak Detection

A Comprehensive Approach to Leak Detection A Comprehensive Approach to Leak Detection Real Water Loss: A Real Issue for Water Utilities Every day, water utilities lose billions of gallons of water designated for public use. In many areas of the

More information

Unit 2 Fire Safety (Zone 2)

Unit 2 Fire Safety (Zone 2) IFE Level 3 Diploma in Fire Safety and Fire Science Unit 2 Fire Safety (Zone 2) Examiner Report March 2017 Introduction 21% of the candidates that sat the examination attained a pass. Candidates did not

More information

Micrel Case Study: Reducing Costs for Industrial Gas Cylinders Management 14 January 2008

Micrel Case Study: Reducing Costs for Industrial Gas Cylinders Management 14 January 2008 Micrel Case Study: Reducing Costs for Industrial Gas Cylinders Management 14 January 2008 CHALLENGE: Micrel owns a 150mm fab which has a capacity of 30,000 wafer starts per month. This fab was constructed

More information

TRISKAIDEKAPHILIA: 2016 GRAPHIC DESIGN SENIOR SHOW. Jordan Colston. Honors College. East Carolina University. In Partial Fulfillment of the

TRISKAIDEKAPHILIA: 2016 GRAPHIC DESIGN SENIOR SHOW. Jordan Colston. Honors College. East Carolina University. In Partial Fulfillment of the TRISKAIDEKAPHILIA: 2016 GRAPHIC DESIGN SENIOR SHOW by Jordan Colston A Senior Honors Project Presented to the Honors College East Carolina University In Partial Fulfillment of the Requirements for Graduation

More information

Enterprise GIS Architecture Deployment Options

Enterprise GIS Architecture Deployment Options 2013 Esri International User Conference July 8 12, 2013 San Diego, California Technical Workshop Enterprise GIS Architecture Deployment Options Andrew Sakowicz Esri UC2013. Technical Workshop op. Introduction

More information

CHAIR AND MEMBERS PLANNING & ENVIRONMENT COMMITTEE JOHN M. FLEMING MANAGING DIRECTOR, PLANNING AND CITY PLANNER

CHAIR AND MEMBERS PLANNING & ENVIRONMENT COMMITTEE JOHN M. FLEMING MANAGING DIRECTOR, PLANNING AND CITY PLANNER TO: FROM: SUBJECT: CHAIR AND MEMBERS PLANNING & ENVIRONMENT COMMITTEE JOHN M. FLEMING MANAGING DIRECTOR, PLANNING AND CITY PLANNER APPLICATION BY: THE CORPORATION OF THE CITY OF LONDON OUTDOOR PATIOS IN

More information

Environmental Footprint ENERGY

Environmental Footprint ENERGY Environmental Footprint ENERGY Learning Environment: Indoor (and outdoor) Prep Time: 10 minutes Length of Lesson: 1 hr Key Vocabulary: Renewable, nonrenewable energy Staffing: 1 adult Materials: All materials

More information

Can Short Term ARRA Stimulus Funding Achieve Long Term Market Transformation? 1

Can Short Term ARRA Stimulus Funding Achieve Long Term Market Transformation? 1 Can Short Term ARRA Stimulus Funding Achieve Long Term Market Transformation? 1 Jane Colby, Cadmus Group, Boulder, CO Scott Davis, Cadmus Group, Boulder, CO ABSTRACT The New York Energy $mart SM Products

More information

Experimental Implementation of Spectrum Sensing Using Cognitive Radio

Experimental Implementation of Spectrum Sensing Using Cognitive Radio 1 Experimental Implementation of Spectrum Sensing Using Cognitive Radio Prepared by: Kendall Schmalz Faculty Advisors: Dr. Yanxiao Zhao Assistant Professor, Electrical and Computer Engineering Dr. Thomas

More information

Heat Exchanger- Lockheed Martin

Heat Exchanger- Lockheed Martin 1 Heat Exchanger- Lockheed Martin Engineering Design 100, Section 22, Dr. Ritter Team 4, MODZ Zeping (Calvin) Cheng (zbc5056@psu.edu), Olivia Connors (ojc5027@psu.edu), Madison Powers (mlp5514@psu.edu),

More information

DEVELOPMENT OF THE TANGO ALARM SYSTEM

DEVELOPMENT OF THE TANGO ALARM SYSTEM 10th ICALEPCS Int. Conf. on Accelerator & Large Expt. Physics Control Systems. Geneva, 10-14 Oct 2005, WE3B.1-70 (2005) DEVELOPMENT OF THE TANGO ALARM SYSTEM L. Pivetta Sincrotrone Trieste, Trieste, Italy

More information

PRODUCT SELECTOR. #LivingspaceAU

PRODUCT SELECTOR. #LivingspaceAU PRODUCT SELECTOR #LivingspaceAU www.thelivingspace.com.au What is Living Space? There s something about natural light that makes us human. It s a warmth that touches us deep inside. Sure, we need the sun

More information

Solutions for Smarter Home Security

Solutions for Smarter Home Security (817) 602-5792 texasappliancehomeautomation.com License # - B06249301 Solutions for Smarter Home Security An Alarm.com powered system provides solutions to monitor and protect your home and everyone in

More information

OpenCL for FPGAs/HPC Case Study in 3D FFT

OpenCL for FPGAs/HPC Case Study in 3D FFT OpenCL for FPGAs/HPC Case Study in 3D FFT Ahmed Sanaullah Martin Herbordt Vipin Sachdeva Boston University Silicon Therapeutics OpenCL for FPGAs/HPC: Case Study in 3D FFT 11/15/2017 What gives FPGAs high

More information

Problem (whatistheproblem) Solution (whatisthesolution) Green Roofs (whatisagreenroof) Solar (PV) and Green Roofs

Problem (whatistheproblem) Solution (whatisthesolution) Green Roofs (whatisagreenroof) Solar (PV) and Green Roofs Problem (whatistheproblem) Solution (whatisthesolution) Green Roofs (whatisagreenroof) Solar (PV) and Green Roofs City Hall Chicago Project Management and Construction: Jörg Breuning, andcharlie Miller

More information

Oral Syringe Product Stability &

Oral Syringe Product Stability & Oral Syringe Product Stability & Print Durability Comparison The Comar Oral Syringe dispenser design has proven to be the most stable product for ease of use and legibility when exposed to a series of

More information

Modeling and Simulation of Axial Fan Using CFD Hemant Kumawat

Modeling and Simulation of Axial Fan Using CFD Hemant Kumawat Modeling and Simulation of Axial Fan Using CFD Hemant Kumawat Abstract Axial flow fans, while incapable of developing high pressures, they are well suitable for handling large volumes of air at relatively

More information

Printing Ink Market in Russia to Market Size, Development, and Forecasts

Printing Ink Market in Russia to Market Size, Development, and Forecasts Printing Ink Market in Russia to 2019 - Market Size, Development, and Forecasts Published: 5/2015 Global Research & Data Services Table of Contents List of Tables Table 1 Demand for printing ink in Russia,

More information

South Meridian. Vision. Action

South Meridian. Vision. Action South Meridian Vision Action Table of Contents Introduction... 1 Goal of the Vision Process...1 This is a really big deal!...1 A foreword thought...2 The Process... 2 Description of the Workshop Game...3

More information

2008 North American Hand Dryer Green Excellence of the Year Award. Dyson B2B Inc.

2008 North American Hand Dryer Green Excellence of the Year Award. Dyson B2B Inc. 2008 North American Hand Dryer Green Excellence of the Year Award Dyson B2B Inc. The 2008 Frost & Sullivan Green Excellence Award in Product Innovation in the North American hand dryer market is presented

More information

Water Heating, Boiler, and Furnace Cost Study (RES 19)

Water Heating, Boiler, and Furnace Cost Study (RES 19) Water Heating, Boiler, and Furnace Cost Study (RES 19) Final Report Prepared for: The Electric and Gas Program Administrators of Massachusetts Part of the Residential Evaluation Program Area Submitted

More information

How to modify Volvo Alarm Siren after it has given Message: Alarm system Service Required Ver. 1.2

How to modify Volvo Alarm Siren after it has given Message: Alarm system Service Required Ver. 1.2 How to modify Volvo Alarm Siren after it has given Message: Alarm system Service Required Ver. 1.2 We will show how to modify the Alarm Siren Unit without using Battery. If you are not sure or do not know

More information

OREM CITY OREM UTOPIA SURVEY 2018

OREM CITY OREM UTOPIA SURVEY 2018 OREM CITY OREM UTOPIA SURVEY 2018 1 Overall quality of life in Orem is high and residents are generally satisfied with the way the city is being run. However, compared to 2014 a somewhat smaller proportion

More information

Efficiency of Non-Azeotropic Refrigerant Cycle

Efficiency of Non-Azeotropic Refrigerant Cycle Purdue University Purdue e-pubs International Refrigeration and Air Conditioning Conference School of Mechanical Engineering 1998 Efficiency of Non-Azeotropic Refrigerant Cycle X. Liu Carrier Corporation

More information

Comparative Study of Technological Trend between DAIKIN and Panasonic in the Field of Air Conditioner

Comparative Study of Technological Trend between DAIKIN and Panasonic in the Field of Air Conditioner Comparative Study of Technological Trend between DAIKIN and Panasonic in the Field of Air Conditioner Nhivuong Pham, Yoshitoshi Tanaka Tokyo Institute of Technology, Management of Technology Dept., Tokyo,

More information

ATTIC FIRES. Part 3 HEAVY SMOKE SHOWING

ATTIC FIRES. Part 3 HEAVY SMOKE SHOWING ATTIC FIRES Part 3 HEAVY SMOKE SHOWING ATTIC FIRE TACTICS HEAVY SMOKE SHOWING Size-Up Investigation Life Priority Follow-Up Changing Tactics Assign Resources Locate The Fire Salvage Fire Attack Overhaul

More information

ENVIRONMENTAL MONITORING IN PREPARATION FOR THE INSTALLATION OF A GREEN ROOF

ENVIRONMENTAL MONITORING IN PREPARATION FOR THE INSTALLATION OF A GREEN ROOF ENVIRONMENTAL MONITORING IN PREPARATION FOR THE INSTALLATION OF A GREEN ROOF Andrew Martin School of Civil Engineering, Purdue University, West Lafayette, Indiana ABSTRACT Green roofs are becoming an increasingly

More information

MODULATING CONTROL OF LOW NO X BURNERS

MODULATING CONTROL OF LOW NO X BURNERS MODULATING CONTROL OF LOW NO X BURNERS ABSTRACT Manufacturers of today s wall-fired low NO X burners (LNB) recommend a single-position secondary air volume control for in-service burners. Air register

More information

Battery Performance Alert: A TOOL FOR IMPROVED PATIENT MANAGEMENT FOR DEVICES UNDER BATTERY ADVISORY

Battery Performance Alert: A TOOL FOR IMPROVED PATIENT MANAGEMENT FOR DEVICES UNDER BATTERY ADVISORY Battery Performance Alert: A TOOL FOR IMPROVED PATIENT MANAGEMENT FOR S UNDER BATTERY ADVISORY VERSION 1.0 AUGUST 8, 2017 Abstract: BACKGROUND: In October 2016, St. Jude Medical issued an advisory on a

More information

Title Page. Report Title: Downhole Power Generation and Wireless Communications. for Intelligent Completions Applications

Title Page. Report Title: Downhole Power Generation and Wireless Communications. for Intelligent Completions Applications Title Page Report Title: Downhole Power Generation and Wireless Communications for Intelligent Completions Applications Type of Report: Quarterly Progress Report Project Report Period: Start Date April

More information

The Emergence of Triboelectric Technology

The Emergence of Triboelectric Technology September 1998 September 1998, Pollution Engineering The Emergence of Triboelectric Technology Triboelectric monitoring, now accepted by the U.S. Environmental Protection Agency, provides continuous monitoring

More information

Management of In-Bin Natural Air Grain Drying Systems to Minimize Energy Cost

Management of In-Bin Natural Air Grain Drying Systems to Minimize Energy Cost Know how. Know now. EC710 Management of In-Bin Natural Air Grain Drying Systems to Minimize Energy Cost Thomas W. Dorn, Extension Educator With prices for all energy sources rising dramatically in recent

More information

General HVAC Recommendations

General HVAC Recommendations General HVAC Recommendations DESIGN GUIDELINES FOR ENERGY EFFICIENT HVAC SYSTEMS Thank you for your interest in energy efficiency! Energy efficient heating, ventilation, and air conditioning (HVAC) equipment

More information

Identification of Aberrant Railroad Wayside WILD and THD Detectors: Using Industry-wide Railroad Data. June 10, 2014

Identification of Aberrant Railroad Wayside WILD and THD Detectors: Using Industry-wide Railroad Data. June 10, 2014 Identification of Aberrant Railroad Wayside WILD and THD Detectors: Using Industry-wide Railroad Data June 10, 2014 1 Agenda Background / Overview Solution Approach Groupings Criteria Sites Speed Weight

More information

Utilization of advanced clutter suppression algorithms for improved standoff detection and identification of radionuclide threats

Utilization of advanced clutter suppression algorithms for improved standoff detection and identification of radionuclide threats Physical Sciences Inc. VG14-048 Utilization of advanced clutter suppression algorithms for improved standoff detection and identification of radionuclide threats Bogdan R. Cosofret, Kirill Shokhirev and

More information

Application of a Reagentless Direct Conductivity Online TOC Analyzer for Clean-in-Place Verification

Application of a Reagentless Direct Conductivity Online TOC Analyzer for Clean-in-Place Verification Application of a Reagentless Direct Conductivity Online TOC Analyzer for Clean-in-Place Verification By Paul Farrell, Emek Blair (Hach Company) and Torben Andersen (McFlusion) ABSTRACT This paper describes

More information

Homework 8: Patent Liability Analysis Due: Thursday, March 24, at Classtime

Homework 8: Patent Liability Analysis Due: Thursday, March 24, at Classtime Homework 8: Patent Liability Analysis Due: Thursday, March 24, at Classtime Team Code Name: G.R.O.V.E.R. Group No. 7 Team Member Completing This Homework: Stan Miller NOTE: This is the second in a series

More information

Performance Comparison of Hydronic Secondary Loop Heat Pump and Conventional Air-Source Heat Pump ABSTRACT 1. INTRODUCTION

Performance Comparison of Hydronic Secondary Loop Heat Pump and Conventional Air-Source Heat Pump ABSTRACT 1. INTRODUCTION 2597, Page 1 Performance Comparison of Hydronic Secondary Loop Heat Pump and Conventional Air-Source Heat Pump Ian H. BELL 1 *, James E. BRAUN 2 1 Bell Thermal Consultants ian.h.bell@gmail.com 2 Purdue

More information

Installation Guide for inbiox60 Series Access Control Panel

Installation Guide for inbiox60 Series Access Control Panel Installation Guide for inbiox60 Series Access Control Panel Version: 1.0 Date: April, 2011 About This Manual This manual is a guide to installation and connection of the inbiox60 series access control

More information

On a GPU Acceleration of the Stochastic Grid Bundling Method

On a GPU Acceleration of the Stochastic Grid Bundling Method On a GPU Acceleration of the Stochastic Grid Bundling Method Álvaro Leitao Rodríguez and Cornelis W. Oosterlee Technische Universiteit Delft (TU Delft) - Centrum Wiskunde & Informatica (CWI) 13/6/2014

More information

Computer Modelling and Simulation of a Smart Water Heater

Computer Modelling and Simulation of a Smart Water Heater Computer Modelling and Simulation of a Smart Water Heater Maria Kathleen Ellul University of Malta ellul_maria@yahoo.com Adrian Muscat University of Malta adrian.muscat@um.edu.mt Abstract A computational

More information

Application of Fuzzy Control in Central Air-conditioning Terminal Units Energy-saving

Application of Fuzzy Control in Central Air-conditioning Terminal Units Energy-saving www.ijcsi.org 123 Application of Fuzzy Control in Central Air-conditioning Terminal Units Energy-saving Yanqiu Zhao 1, Yingli Lv 1, Yong Gu 1, Shucheng Liu 2 1 Department of Electrical Engineer, Hebei

More information

Testing of the IBM z-series Modular Refrigeration Unit. Terrence Quinn IBM Integrated Supply Chain Procurement Engineering

Testing of the IBM z-series Modular Refrigeration Unit. Terrence Quinn IBM Integrated Supply Chain Procurement Engineering Testing of the IBM z-series Modular Refrigeration Unit Terrence Quinn IBM Integrated Supply Chain Procurement Engineering Overview: This presentation will describe the methodology and hardware used in

More information

Achieving Functional Safety using Time-Triggered Architectures

Achieving Functional Safety using Time-Triggered Architectures - 1 - Achieving Functional Safety using Time-Triggered Architectures Michael J. Pont, PhD SafeTTy Systems Ltd Vector UK Conference 2014 Functional Safety 18 June 2014 Abstract This talk will explore some

More information

Real-Time Root Cause Analysis for Complex Technical Systems

Real-Time Root Cause Analysis for Complex Technical Systems Real-Time Root Cause Analysis for Complex Technical Systems Jan Eric Larsson, Joseph DeBor IEEE HPRCT Monterey, CA, August 29 th 2007 Handling Alarm Cascades Root cause analysis: find the initiating events

More information

Sensor Technology. Summer School: Advanced Microsystems Technologies for Sensor Applications

Sensor Technology. Summer School: Advanced Microsystems Technologies for Sensor Applications Sensor Technology Summer School: Universidade Federal do Rio Grande do Sul (UFRGS) Porto Alegre, Brazil July 12 th 31 st, 2009 1 Outline of the Lecture: Philosophy of Sensing Thermal Sensors Chemical Sensors

More information

Experion LX Direct Station Specification

Experion LX Direct Station Specification Technical Information Experion LX Direct Station Specification LX03-210-110 Release 110 January 2014, Version 1 Experion LX Station Specification, LX03-210-110 Revision History Revision Date Description

More information

Modelling Appliance Loads in Multi-Unit Residential Buildings

Modelling Appliance Loads in Multi-Unit Residential Buildings Modelling Appliance Loads in Multi-Unit Residential Buildings April 2018 Prepared by: Report For: Contents 1. Introduction... 3 2. Resource Review... 5 3. Recommendations... 7 3.1 What is the reference...

More information

TSI AEROTRAK PORTABLE PARTICLE COUNTER MODEL 9110

TSI AEROTRAK PORTABLE PARTICLE COUNTER MODEL 9110 TSI AEROTRAK PORTABLE PARTICLE COUNTER MODEL 9110 APPLICATION NOTE CC-107 Introduction This purpose of this document is to detail the advanced, state of the art features TSI has incorporated in the design

More information

Water Mist-Based Fire Suppression Modelling of an Office Space Scenario

Water Mist-Based Fire Suppression Modelling of an Office Space Scenario Water Mist-Based Fire Suppression Modelling of an Office Space Scenario Hai Jiang, Vaidya Sankaran, Med Colket, May Corn United Technologies Research Center Kati Laakkonen Marioff Corporation 14th International

More information

FEATURE ARTICLE. The PLCA-800 Series Inline Particle Sensor for Chemical Solutions. Yoshihito Yuhara. Abstract

FEATURE ARTICLE. The PLCA-800 Series Inline Particle Sensor for Chemical Solutions. Yoshihito Yuhara. Abstract FEATURE ARTICLE Technical Reports The PLCA-800 Series Inline Particle Sensor for Chemical Solutions Yoshihito Yuhara Front end process Sensor module Wafer preparation process Film formation Photo lithography

More information

Workshop on AFDD for RTUs Moving from R&D to Commercialization July 13, Introduction

Workshop on AFDD for RTUs Moving from R&D to Commercialization July 13, Introduction Workshop on AFDD for RTUs Moving from R&D to Commercialization July 13, 2014 Introduction Jim Braun Ray W. Herrick Laboratories Purdue University West Lafayette, IN 47907 Slide 1 Acknowledgement This workshop

More information

Hot Water System Performance Testing at PG&E s Applied Technology Services: Mimicking a Full-Service Restaurant in the Laboratory

Hot Water System Performance Testing at PG&E s Applied Technology Services: Mimicking a Full-Service Restaurant in the Laboratory Hot Water System Performance Testing at PG&E s Applied Technology Services: Mimicking a Full-Service Restaurant in the Laboratory Amin Delagah Project Engineer Fisher Nickel, a division of Frontier Energy

More information

New fire control system design for fire fighting and rescue equipment Jianguo LUO1, a, Zehao BU2, b

New fire control system design for fire fighting and rescue equipment Jianguo LUO1, a, Zehao BU2, b 6th International Conference on Information Engineering for Mechanics and Materials (ICIMM 2016) New fire control system design for fire fighting and rescue equipment Jianguo LUO1, a, Zehao BU2, b 1 Department

More information

Section 9. Comparing Energy Consumption: More for Your Money. What Do You See? What Do You Think? Investigate. Learning Outcomes

Section 9. Comparing Energy Consumption: More for Your Money. What Do You See? What Do You Think? Investigate. Learning Outcomes Section 9 Comparing Energy Consumption: More for Your Money Section 9 Comparing Energy Consumption: More for Your Money What Do You See? Learning Outcomes In this section, you will Measure and compare

More information

CCTV, HOW DOES IT WORK?

CCTV, HOW DOES IT WORK? August 25, 2005 CCTV FLAME DETECTION Video flame detection is an attractive concept. We ve all heard the adage, seeing is believing. In short, we trust our eyes to deliver accurate information. New designs

More information

Eliminating Condensation and Resulting Corrosion. Sulphur Tail Gas and De-Gas Lines

Eliminating Condensation and Resulting Corrosion. Sulphur Tail Gas and De-Gas Lines Eliminating Condensation and Resulting Corrosion In Sulphur Tail Gas and De-Gas Lines By David R. Hornbaker, Controls Southeast, Inc. and Graham M. MacPherson, P.Eng., Millenia Resource Consulting Abstract

More information

Moving to the Cloud: The Potential of Hosted Central Station Services

Moving to the Cloud: The Potential of Hosted Central Station Services Moving to the Cloud: The Potential of Hosted Central Station Services Moving to the Cloud: The Potential of Hosted Central Station Services The global market for alarm monitoring services is expected to

More information

TEXAS CHAPTER AMERICAN SOCIETY OF LANDSCAPE ARCHITECTS 2019 PROFESSIONAL AWARDS GUIDELINES

TEXAS CHAPTER AMERICAN SOCIETY OF LANDSCAPE ARCHITECTS 2019 PROFESSIONAL AWARDS GUIDELINES TEXAS CHAPTER AMERICAN SOCIETY OF LANDSCAPE ARCHITECTS 2019 PROFESSIONAL AWARDS GUIDELINES The Texas Chapter of the American Society of Landscape Architects invites, from the Chapter membership, submission

More information