資訊工程學系謝孫源講座教授特別邀請來自【美國喬治華盛頓大學Ahmed Louri教授】及【美國波士頓東北大學Fabrizio Lombardi教授】親臨本校演講，誠摯邀請您一同參與交流，演講的主題與資訊如下：
報名網址：https://reurl.cc/vn1EZN

--------------------------------------------------------------------------------------------------------------------------------------
Date：108/09/19(四) 5:00 PM- 6:00 PM
Venue:國立成功大學 @光復校區/國際會議廳
Talk Title: Power-Efficient and Reliable Network-on-Chips for Scalable Multicore Architectures
Speaker:美國喬治華盛頓大學Ahmed Louri教授
Abstract:
Today’s microprocessor designers have resorted to increasing the number of cores per die as a power-efficient approach to performance improvement, leading to the Chip Multiprocessors (CMPs) or the multicore era. Both industrial and academic roadmaps project that hundreds to thousands of general-purpose tera-op CMPs will be needed within a decade in order to satisfy future needs for high-performance computing.

The proliferation of multiple cores on the same die heralded the advent of a communication-centric, rather than computation-centric systems, to the forefront of computing design wherein the design of the on-chip interconnect fabric connecting various modules, namely the processing cores, cache banks, and memory units, has become extremely important. Further, the continuing scaling of technology has accentuated the on-chip interconnect problem since global wire delays do not scale down as fast as gate delays. To address the growing wire delay problems and improve the performance, a growing number of multicore designs have adopted a more flexible and scalable packet-switched architecture called Network-on-Chip (NoC) architectures. As the number of cores on the chip keeps growing to satisfy power and performance scaling, the NoC design has become the most critical element to achieving the performance potential of future CMPs. Among the challenges facing current NoC design, power dissipation and reliability have been identified as the most critical ones.

In this talk, I will first discuss several research challenges facing multicore architectures and NoC design. Next, I will present some of our ongoing efforts to address these issues using architectural innovations as well as a synergistic mix of emerging interconnect technologies.



Date：108/09/20(五) 10:00 AM -11:00 AM
Venue:國立成功大學 @光復校區國際會議廳
Talk Title: Machine Learning Role in the Design of Efficient Computer Architectures and Systems of the Future.
Speaker:美國喬治華盛頓大學Ahmed Louri教授
Abstract:

While the computing world is reaching the limits of Moore’s Law, demands for performance in excess of 1 million trillion floating-point operations per second (1 exaflops) are arising from novel paradigms to address applications in big data, machine learning and analytics from nearly countless resources. This computing challenge has never been as complex or as critically important as it is now. We are witnessing a new computing era calling for a migration from an algorithmic based compute world to a machine-learning based, data-intensive, memory-centric computing paradigm in which human capabilities are scaled and magnified. This paradigm shift is driven by the abundance of data (big data era) and the computing resources required to process it in application spaces spanning mobile/embedded computing, cloud/edge computing, and the internet of things. This new trend has reinvigorated computer architecture research. Researchers are considering new ways forward based in machine learning.

In this talk, I will first briefly describe the new mega trends in computer architecture research. I will then present our own research efforts and introduce a design paradigm that exploits the use of machine-learning to simultaneously improve power-efficiency, performance, and reliability for heterogeneous multicore architectures and on-chip communications. The aim is to provide insights and directions of integrating architectural innovations with machine learning in a holistic approach for future computer architecture research.


【講者介紹】:
來自美國的學者Ahmed Louri教授，自2015年8月起任教於美國喬治華盛頓大學電機系（Department of Electrical and Computing Engineering, George Washington University, Washington DC），並為David and Marilyn Karlgaard授予講座教授職位的持有人（the David and Marilyn Karlgaard Endowed Chair Professor）。2015年至2016年期間，同時擔任美國喬治華盛頓大學電機系主席一職。1988年至2015年，於美國亞利桑那大學電子與計算機工程系擔任教授。

Ahmed Louri教授是位世界知名的學者，他是IEEE Fellow，也多次獲得歐亞各國所頒發的榮譽獎項，Ahmed Louri教授是IEEE Transactions on Computers（2019-2023）的總編輯，此期刊是IEEE計算機學會的旗艦期刊。他同時是IEEE International Conference on High Performance Computing and Communications (HPCC-2019)的主席，也是IEEE Transactions on Emerging Technologies for Computing(2015~) 和IEEE Transaction on Sustainable Computing(2016~)的副編輯，並被任命為2019年the IEEE CS Fellow Evaluation Committee的主席。他發表了160多篇期刊論文和同行評審的會議論文，並且是多項美國和國際專利的共同發明人。並經常受邀擔任研討會的主講者，同時也擔任眾多國際會議的執行和技術計劃委員會成員，他也擔任資助機構的審查員和小組成員，包括NSF，Agence Nationale de la Recherche (France), Qatar National Research Foundation, Hong Kong National Research Council, European Research Council等等。

--------------------------------------------------------------------------------------------------------------------------------------
Date：108/09/19(四) 4:00 PM - 5:00 PM
Venue: 國立成功大學 @光復校區國際會議廳
Talk Title: Room Temperature Magnetic Field Sensing Using the Magneto-Electric Effect in Single Phase Hexaferrite Thin
Speaker:美國波士頓東北大學Fabrizio Lombardi教授
Abstract：

The development of new magnetic field sensors that are capable of detecting low-frequency small magnetic fields at room temperature has long been an important area of research. The ability to detect small fields with high sensitivity is very critical for a wide range of uses including non-invasive biomedical applications such as magnetocardiography and magnetoencepthalogy. Novel techniques for magnetic field sensing applications based on the magnetoelectric (ME) effect have previously been explored in bulk SrCo2Ti2Fe8O19 materials and single-phase hexaferrite thin films. Since the Magneto-Electric (ME) effect can be used for passive magnetic field sensing at room temperature, it is highly suitable for designing the next generation of magnetic field sensors. Initially, we have observed that in single-phase hexaferrite materials, a thin layer of ME hexaferrite film can be used to convert an input magnetic field into an electric voltage while producing very high ME coupling at room temperature.

Based on these observations, an alternative method for detecting both DC and low frequency AC magnetic fields using hexaferrite thin films will be presented in this talk. The proposed sensors are designed, fabricated and assessed to have a good linear operating range, while consuming only a small amount of power and occupying a relatively small chip area. The design and experimental implementation of bi-directional ME sensors that can determine both magnitude and direction of an incident magnetic field as an alternative to portable magnetic field sensors such as fluxgate magnetometers are also briefly presented. Finally, this presentation will focus on modeling, fabricating, testing and evaluating ME-based sensors in terms of their detection range, noise performance, direction sensing capabilities as well as compatibility with existing CMOS technologies.



Date：108/09/20(五) 11:00 AM- 12:00 PM
Venue:國立成功大學 @光復校區國際會議廳
TalkTitle: New Architectural Advances in Approximate Computing
Speaker:美國波士頓東北大學Fabrizio Lombardi教授
Abstract：
This presentation deals with the new computational paradigm of approximate (inexact) computing at architecture level. While inexact computing introduces inaccuracy in the computed results, it has been advocated as a possible alternative in the nanoscales to improve computational speed, reduce power consumption and hardware complexity. Approximate computing is application dependent and therefore it can be used only in specific design spaces for an efficient implementation.

In this talk, two opportunities presented by approximate computing are considered with respect to two design spaces: fast Fourier transform and voting. In the first case, bit-width of operands is analyzed under two different constrains, namely area and frequency dominated scenarios. In both cases, the Single-path Delay Feedback (SDF) for pipeline implementation is considered. For voting, the inexact design of a duplex scheme is presented for reliable computing. The implications of this “soft” technique are then outlined.


【講者簡介】:

來自美國的學者 Fabrizio Lombardi教授，自1998年至今就任教於美國波士頓東北大學電機系（Department of Electrical and Computing Engineering, Northeastern University, Boston），並且為International Test Conference/ITC授予教授職位的持有人。1998至2004年期間，他並同時擔任美國波士頓東北大學電機系系主任一職。

Fabrizio Lombardi教授是位世界著名的學者，他是電機電子工程師學會的院士（Institute ofElectrical andElectronicsEngineers，簡稱IEEE），也是IEEE Transactions onNanotechnology的主編輯，同時也是IEEE Nanotechnology Council的執行委員會成員。他被任命為許多非營利組織的執行委員，例如Computing-in-the-Core, now code.org the non-partisan advocacy coalition for K-12 Computer Science education as well as the Computer Society當選為兩屆理事會任期成員（2012-2017）和IEEE Future Directions Committee的任命成員（2014-2017）。

Fabrizio Lombardi教授多次獲得由各國頒發的專業獎項：加拿大維多利亞大學British Columbia Advanced System Institute頒發的 Visiting Fellowship（1988）、兩次獲得Texas Experimental Engineering Station頒發的Research Fellowship（1991-1992,1997-1998）、Texas Experimental Engineering Station頒發的the Halliburton Professorship（1995）、 東北大學傑出工程研究獎（2004）、日本科學教育部所頒發的國際研究獎（1993-1999）及IEEE Transactions on Engineering Foundation 1985/86研究啟動獎和摩托羅拉奧斯汀銀獎（1996）。

Fabrizio Lombardi教授是IEEE Transactions on Computers（1996-2000）的副編輯和IEEE-CS的傑出訪問者（1990-1993）。自2000年以來，他一直擔任IEEE Transactions on Computers的副主編。目前，他還是IEEE Design and Test Magazine的副編輯和IEEE-CS的傑出訪問者; 他還擔任IEEE Test Technology Technical Council 項下的Nanotechnology Devices and Systems委員會主席。Fabrizio Lombardi教授參與組織了許多由專業組織主辦的國際研討會以及特刊的特邀編輯，這些文章包括IEEE Transactions on Computers、IEEE Transactions on Instrumentation and Measurement、IEEE Micro Magazine和IEEE Design and Test Magazine。他是IEEE Symposium on Network Computing and Applications的創始主席。他在這些領域廣泛出版並合著/編輯了十本書。

Fabrizio Lombardi教授在Electronic and Electrical Engineering的領域中占有一席之地，是位相當傑出的研究學者，其研究主軸包含數位系統的測試與設計（Testing and design of digital systems）、量子和納米計算（quantum and nano computing）、自動測試設備系統(ATE/Automatic test equipment Systems）、可配置網路計算（configurable /network computing）、缺陷容差（defect tolerance）和計算機輔助設計超大型積體電路 (CAD VLSI)。