Parallel systems that support the shared memory abstraction are becoming widely accepted in many areas of computing. Writing correct and efficient programs for such systems requires a formal specification of memory semantics, called a memory consistency model. The most intuitive model—sequential consistency—greatly restricts the use of many performance optimizations commonly used by uniprocessor hardware and compiler designers, thereby reducing the benefit of using a multiprocessor. To alleviate this problem, many current multiprocessors support more relaxed consistency models. Unfortunately, the models supported by various systems differ from each other in subtle yet important ways. Furthermore, precisely defining the semantics of each model often leads to complex specifications that are difficult to understand for typical users and builders of computer systems. The purpose of this tutorial paper is to describe issues related to memory consistency models in a way that would be understandable to most computer professionals. We focus on consistency models proposed for hardware-based shared-memory systems. Many of these models are originally specified with an emphasis on the system optimizations they allow. We retain the system-centric emphasis, but use uniform and simple terminology to describe the different models. We also briefly discuss an alternate programmer-centric view that describes the models in terms of program behavior rather than specific system optimizations. 1

Digestor is a software system which automatically re-authors arbitrary documents from the World-Wide Web to display appropriately on small screen devices such as PDAs and cellular phones, providing device-independent access to the Web. Digestor is implemented as an HTTP proxy which dynamically re-authors requested Web pages using a heuristic planning algorithm and a set of structural page transformations to achieve the best looking document for a given display size. 1. Introduction Access to World-Wide Web documents from personal electronic devices has been demonstrated in research projects [2,10,17,18], and is now becoming a commercial reality. General Magic's Presto!Links for Sony's MagicLink, AllPen's NetHopper for the Newton and Sharp's MI-10 (Figure 1, shown at right), all provide WWW browsers for PDA class devices, while the Nokia 9000 Communicator and Samsung's Duett provide Web access capabilities from cellular phones. Figure 1. Digestor: Device-Independent Access to the W...

research relevant to the design and application of high performance scientific computers. We test our ideas by designing, building, and using real systems. The systems we build are research prototypes; they are not intended to become products. There are two other research laboratories located in Palo Alto, the Network Systems

7. References Low Bandwidth Surfing in a High Bandwidth World Today's WWW is beginning to burst at the seams due to lack of bandwidth from servers to clients. Most WWW pages are designed with high-bandwidth users in mind (i.e. 10 Mbit Ethernet) , yet a large percentage of web clients are run over low-speed 28.8 or 14.4 modems, and users are increasingly considering wireless services such as cellular modems running at 4800-9600 baud. A recent study by a popular server of shareware, Jumbo, revealed that about 1 in 5 users were connecting with graphics turned off, to eliminate the annoying latency of loading web pages. This is of particular concern to corporate advertisers, who are paying for the visibility of their corporate logo. Since there is no way to optimize a single page for delivery to both high-bandwidth and low-bandwidth clients, some sites, such as [Xinside] , offer multiple versions of pages (no graphics, minimal graphics, full graphics). Most sites, however

mogul @ wrl.dec.com The Internet has experienced exponential growth in the use of the World-Wide Web, and rapid growth in the use of other Internet services such as VSENET news and electronic mail. These applications qualitatively differ from other network applications in the stresses they impose on busy server systems. Unlike traditional distributed systems, Internet servers must cope with huge user communities, short interactions, and long network latencies. Such servers require different kinds of operating system features to manage their resources effectively. 1

In this paper we describe the Mobile Link (m-Links) infrastructure for utilizing existing World Wide Web content and services on wireless phones and other very small Internet terminals. Very small devices, typically with 3-20 lines of text, provide portability and other functionality while sacrificing usability as Internet terminals. In order to provide access on such limited hardware we propose a small device web navigation model that is more appropriate than the desktop computer's web browsing model. We introduce a middleware proxy, the Navigation Engine, to facilitate the navigation model by concisely displaying the Web's link (i.e., URL) structure. Because not all Web information is appropriately "linked," the Navigation Engine incorporates data-detectors to extract bits of useful information such as phone numbers and addresses. In order to maximize program-data composibility, multiple network-based services (similar to browser plug-ins) are keyed to a link's attributes such as its MIME type. We have built this system with an emphasis on user extensibility and we describe the design and implementation as well as a basic set of middleware services that we have found to be particularly important. Keywords Wireless, wireless web, web phones, middleware, proxy. 1.

research relevant to the design and application of high performance scientific computers. We test our ideas by designing, building, and using real systems. The systems we build are research prototypes; they are not intended to become products. There are two other research laboratories located in Palo Alto, the Network Systems

We propose a progressively reliable transport protocol for delivery of delay-sensitive multimedia over Internet connections with wireless access links. The protocol, termed "Leaky" ARQ, initially permits corrupt packets to be leaked to the receiving application and then uses retransmissions to progressively refine the quality of subsequent packet versions. A Web server would employ Leaky ARQ to quickly deliver a possibly corrupt first version of an image over a noisy bandlimited wireless link for immediate display by a Web browser. Later, Leaky ARQ's retransmissions would enable the browser to eventually display a cleaner image. Forwarding and displaying corrupt error-tolerant image data: (1) lowers the perceptual delay compared to fully reliable packet delivery, and (2) can be shown to produce images with lower distortion than aggressively compressed images when the delay budget only permits weak forward error correction. Leaky ARQ supports delaying of re-transmissions so that initial packet transmissions can be expedited, and cancelling of retransmissions associated with "out-of-date" data. Leaky ARQ can be parameterized to partially retransmit audio and video. We propose to implement Leaky ARQ by modifying Type-II Hybrid/"code combining" ARQ.

To meet the service demands created by the Internet's exponential growth, operators are scrambling to deploy application-level services, including Web caches, commerce servers, and intelligent transformation proxies for mobile "thin clients." On the one hand, the Internet's growth rate places unprecedented scalability and robustness demands on these services; on the other hand, that same growth rate demands that new services be developed, deployed, and evolved at a pace that is precipitous even by the standards of today's desktop software development cycles. We demonstrate that for a certain class of applications, these apparently conflicting goals can be reconciled by completely separating the application logic from the runtime support for scalability and high ...

This paper presents a design capture system in which schematics are translated into a procedural netlist specification language. The circuit designer draws schematics with a standard structured graphics editor that knows nothing about netlists or schematics. The translator program analyzes the structured graphics output file and translates it into a procedural netlist specification. d i g i t a l Western Research Laboratory 250 University Avenue Palo Alto, California 94301 USA ii Table of Contents 1. Introduction 1 2. Basics 2 2.1. Simple Example 2 2.2. Structured Graphics 3 3. Generating Procedures 4 3.1. Frames and Evaluation 4 3.2. 2D Ordering 5 4. Drawing Interpretation 7 4.1. Icons 8 5. Analysis of Non-Evaluation Objects 9 5.1. Binding Text to Objects 9 5.2. Wires 10 5.3. Wire Subscripting 11 6. Error Reporting 11 7. Experiences 12 Acknowledgements 12 References 12 iii iv List of Figures Figure 1: Code Generated for "CELL: orN" 2 Figure 2: 2D ordering of objects 5 Figur...