📄 http:^^www.cs.wisc.edu^~glew^generic-phd-research-interests.html
字号:
Date: Tue, 05 Nov 1996 21:15:18 GMTServer: NCSA/1.5Content-type: text/htmlLast-modified: Thu, 12 Sep 1996 23:37:54 GMTContent-length: 16868<!DOCTYPE HTML PUBLIC><HTML> <HEAD> <title> Andy Glew - Statement of Purpose: Ph.D. Application, 1995 </title> </HEAD> <BODY> <H1>Andy Glew: Statement of Purpose</H1> <H2>(0) What this document is</H2> This document is my <I>Statement of Purpose</I>. It is being prepared as part of my application for admission to Ph.D. studies in Computer Architecture (Electrical/Computer Engineering). <H2>(1) Introduction</H2> My "goals for graduate study and a professional career" can be briefly described as to do leading edge research in high performance, low cost computing design. <p> In section (3) below I discuss these goals in more detail, and also talk about my career options as a computer architect in industry, and as a researcher in industry and academia. <p> In section (4) I discuss my research interests: in section (4.1) at a very high level indicating the true breadth of my interests, but in section (4.2) at a very specific level that I propose for my Ph.D. research. <p> I am almost embarrassed to admit to the degree of abstraction and the theoretical nature of some of the things I discuss in section (4.1). I mention them mainly to show that I am not a boring drone with a one-track mind, incapable of understanding or contributing to fields other than the one in which I am specializing. <p> In section (4.2) I propose a particular plan of research. I have great hope that it will be productive. <p> However, since it is necessary to balance the <I>abstract ideas</I> of my research interests with the <I>practical details</I> of graduate school and my professional career, in section (2) I provide a brief autobiographical sketch, in the hope that this background may help explain why I wish to achieve some of the things I am setting out to do, and give credence to my likelihood of achieving them. <H2>(2) Autobiographical Sketch</H2> I graduated from McGill University with a B.Eng. in Electrical Engineering in 1985. Between 1985 and 1989, I worked as a programmer at a computer graphics company (Formic) and in an operating systems development group (Gould), and as a performance analysis in a company that manufactured computers (Motorola). <p> I completed my M.S. at the University of Illinois, under Professor Wen-Mei Hwu, in 1991. Although my Master's Thesis was on a topic in multiprocessor cache protocols, the bulk of my research was in out-of-order, speculative, CPU microarchitectures. <p> Starting work at Intel in 1991, I got the chance to apply my CPU microarchitecture research, which I had actually begun while an undergraduate at McGill, and carried through the University of Illinois. I was one of the five principal architects of the P6 (now called Pentium Pro) microprocessor. P6 is one of the first, and still most aggressive, out-of-order, speculative microprocessors. <p> Now that the P6 processor project is completed, I have decided to complete the Ph.D. research which I suspended five years ago. I am also entering a new role at Intel, as the first member of the Microcomputer Research Labs' Intel Architecture group. Although competing my Ph.D. necessarily takes priority, I hope to perform a balancing act between academic research and industrial research. <H2>(3) Goals for Graduate Study and a Professional Career</H2> My goals for graduate school are as follows: I have a specific area in mind, research into advanced CPU microarchitectures. I hope that my Ph.D. research will be as relevant as my earlier work, and that the designs I will investigate will be practical enough to influence the design of computers in industry at the completion of my Ph.D. I.e. I hope that my research will influence the design of computers shipping in the years 2005-2015. But I am open to other topics as described in my interests section, section (4.1), below. <p>It should be obvious from my autobiographical sketch, section (2), that I am already well established in my professional career. I have occasionally been accused of being overly theoretical or academic in my analyses of problems in computer design, so I have been sensitive to the dictum "Those who can, do. Those who can't teach (or do research)" but I think that my successful role in P6 disproves this. <p>Long term, I wish to continue to do leading edge work in computer design: initially in industry, but also perhaps in an academic setting. If I stay with Intel my career goal is to become an Intel Fellow, with technical seniority sufficient to influence the direction of the corporation. I do not, however, have plans to become exclusively a manager. On the other hand, I am considering returning to a university to teach and do research, perhaps in my native country, Canada, always with the hope of keeping my industrial contacts to increase the chances of performing relevant research. <H2>(4) Research Interests</H2> In section (4.2) I describe one particular area in which I am interested in doing research. Before this, though, in section (4.1) I describe some very broad areas that are of interest to me. But, again, before this, I describe my motivation most succinctly in section (4.0). <p> It can be seen that, although I have chosen a particular, very practical, area of applied and experimental research in computer architecture, my interests are more general. <H3>(4.0) Motivation</H3> <UL> <LI>I want to make computers faster. <LI>I want to make computers faster, to relieve humans of intellectual drudgery. <LI>I want to make inexpensive computers faster, fast enough to support natural modes of interaction with users. <UL> <LI>I want to make inexpensive computers fast enough so that non-computer literate people like my parents can use them. </UL> <LI>I want to make computers faster, fast enough that "artificial intelligence" can reasonably be studied without the constraints of insufficient computing resources. <LI>I think that the study of computers produces insights into both intelligence and the structures of mathematical and logical systems; for the moment, my contribution is to make computers faster, but I am eventually interested in working in these other areas of knowledge. </UL> <H3>(4.1) Broad, High Level</H3> Overall, I am interested in the augmentation of human intelligence using computers. I am interested enough that, did I not already have a topic in mind, I would consider doing applied research developing software such as "thinking tools", agents, and improved ways of representing human knowledge such as the as-yet unachieved ideal of Vannevar Bush's MEMEX hypertext system. <p> It became obvious to me when I first encountered such issues that the user interface was a major obstacle to truly creating tools that augment human intelligence. Hence my interest in 3D graphics, virtual reality, and speech and handwriting recognition. <p> In the mid-1980s, when I started my career, it was obvious that one of the greatest obstacles to such improved user interfaces and thinking tools was lack of computational power. Therefore, I have spent most of my career, to date, improving computer systems performance, so that one day the user interfaces that I wish to use will be cheap and ubiquitous. <p> Similarly, the fact that most computer systems do not provide real-time responsiveness, even when computationally capable of it, is an obstacle to human interaction with computers. Hence my involvement in real-time operating systems design when I was a software developer, and my provision of the hardware prerequisites for providing "real-timeliness" as a hardware developer. I remain interested in the incorporation of time as a criterion⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -