2008 5th IEEE International Conference on Group IV Photonics, 2008
We introduce a silicon photonic WDM point-to-point network enabled by novel optical proximity com... more We introduce a silicon photonic WDM point-to-point network enabled by novel optical proximity communications. This strictly non-blocking network provides scalable interconnectivity between chips low latency and high bisection bandwidth.
Silicon photonics is a promising technology to scale offchip bandwidth in a power-efficient manne... more Silicon photonics is a promising technology to scale offchip bandwidth in a power-efficient manner. Given equivalent bandwidth, the flexibility of switched networks often leads to the assumption that they deliver greater performance than point-to-point networks on message passing applications with low-radix traffic patterns. However, when optical losses are considered and total optical power is constrained, this assumption no longer holds.
A simulation engine is the collection of components, features and support functions which are cru... more A simulation engine is the collection of components, features and support functions which are crucial to the implementation of an efficient discrete event simulation model. Furthermore, this model can be embedded in a larger application. A good simulation engine combines efficiency, functionality and completeness to enable a model builder to construct a comprehensive, customized model of either a specific system or a class of systems. This paper presents CSIM18, a simulation engine which supports development of applications with efficient, embedded simulation models on a variety of system plat
ABSTRACT Energy-efficiency has become a critical parameter in the design of high-performance comp... more ABSTRACT Energy-efficiency has become a critical parameter in the design of high-performance computing systems. Typically, compute elements consume the most energy in current systems, with memories and interconnect networks close behind. This paper proposes an energy-efficient, high-connectivity, petascale computing system for the year 2020 timeframe by addressing the energy requirements of these three components. We start with projections based on purely evolutionary computer system design trends, then include the impact of breakthroughs in processor design, memory packaging and optical interconnect technologies. Based on these projections, we motivate the development for a 1-pJ/b optical intrasystem interconnect technology that significantly increases system interconnect bandwidth and relieves the distance-based energy dependence of electrical alternatives. We show that improvements in compute, memory, and IO, when simultaneously applied, become a vision for many-chip photonically-interconnected modules that could lead to an order of magnitude improvement in energy efficiency in the 2020 timeframe. The vision hinges on a high-density, energy-efficient optical link that can connect electronic compute and memory elements across short chip-to-chip distances while also capable of kilometer or longer spans across data centers. We discuss the power budget to enable such a link and review experimental progress toward creating an ultra-dense, hybrid-integrated low-power silicon photonic link that will enable this vision.
Proceedings of the fourth symposium on Operating system principles - SOSP '73, 1973
This paper presents a description and analysis of a task scheduling algorithm which is applicable... more This paper presents a description and analysis of a task scheduling algorithm which is applicable to third generation computer systems. The analysis is carried out using a model of a computer system having several identical task processors and a fixed amount of memory. The algorithm schedules tasks having different processor-time and memory requirements. The goal of the algorithm is to
WSC'99. 1999 Winter Simulation Conference Proceedings. 'Simulation - A Bridge to the Future' (Cat. No.99CH37038), 1999
ABSTRACT Developing a behavioral model of a system prior to implementing that system provides sig... more ABSTRACT Developing a behavioral model of a system prior to implementing that system provides significant benefits. In some cases, errors in the design can be detected and corrected prior to implementation. In other cases, having a means for predicting the performance of the system prior to its deployment may aid in the design, implementation and configuration of the system. This paper presents a case for developing simulation models of systems early in the design and implementation process. A number of examples illustrate the benefits which can accrue. 1 INTRODUCTION New systems are becoming increasingly more complex, offering more functionality, greater speed, and higher reliability, all at reduced costs. Along with these new requirements, there is also the fact that market windows are decreasing, time-to-market is decreasing, and the number of product cycles is increasing All of these factors mean that it is more important than ever that all facets of a newly designed system be validated a...
Proceeding of the 2001 Winter Simulation Conference (Cat. No.01CH37304), 2001
CSIM19 is the latest version of the system modeling toolkit from Mesquite Software. CSIM19 offers... more CSIM19 is the latest version of the system modeling toolkit from Mesquite Software. CSIM19 offers many features that enable a modeler to develop robust and realistic models of complex systems. These models represent the system as a collection of processes and resources; in most cases the processes mimic the behavior of the entities of the system as they compete for use of the system's resources. The paper presents CSIM19 and the new features in CSIM19. These features include new capabilities for managing processes at facilities, enhanced handling of messages at mailboxes, and performance improvements. Using the OptQuest optimization package with CSIM19 is also discussed. The article concludes with an extended example
Proceedings of the ACM annual conference on - ACM '72, 1972
This paper describes an experimental study of the performance of a large multi-programmed compute... more This paper describes an experimental study of the performance of a large multi-programmed computer system (the UT-l/CDC 6600 system at the University of Texas at Austin) under systematic variation of available resources and resource allocation algorithms. The experiments were carried out in a controlled and reproducible environment provided by a synthetic job stream generator. The experimental data was recorded by an event-driven software monitor which recorded a complete trace of system activities at the level of system defined events. The study relates resource utilization and queueing patterns to the metric of job completion rate. The experiments undertaken in these studies are single factor experiments. Compensatory reactions by this complex system to variation of individual resources are nonetheless revealed. The experiments also demonstrate the criticality of optimal scheduling of bottleneck resources and offer comparisons of the performance of multi-drive disk units under different conditions of availability and space assignment. The data gathering facility was also run on the production environment to determine base lines for comparison to the experiments as well as for an understanding of the production mode of operation of the system.
ABSTRACT In many cases, a simulation model of a system is used to evaluate alternative configurat... more ABSTRACT In many cases, a simulation model of a system is used to evaluate alternative configurations of that system, with the goal of finding the configuration which maximizes (or minimizes) the value of an objective while meeting all of the stated requirements. The CSIM18/OptQuest package automates this kind of search for the best configuration by combining a powerful simulation engine, CSIM18, and a state-of-the-art optimization package, OptQuest. This paper describes this integrated package for doing optimization and simulation. The paper concludes with two examples: finding the best configuration for a jobshop, and finding the best configuration for a web server.
2008 5th IEEE International Conference on Group IV Photonics, 2008
We introduce a silicon photonic WDM point-to-point network enabled by novel optical proximity com... more We introduce a silicon photonic WDM point-to-point network enabled by novel optical proximity communications. This strictly non-blocking network provides scalable interconnectivity between chips low latency and high bisection bandwidth.
Silicon photonics is a promising technology to scale offchip bandwidth in a power-efficient manne... more Silicon photonics is a promising technology to scale offchip bandwidth in a power-efficient manner. Given equivalent bandwidth, the flexibility of switched networks often leads to the assumption that they deliver greater performance than point-to-point networks on message passing applications with low-radix traffic patterns. However, when optical losses are considered and total optical power is constrained, this assumption no longer holds.
A simulation engine is the collection of components, features and support functions which are cru... more A simulation engine is the collection of components, features and support functions which are crucial to the implementation of an efficient discrete event simulation model. Furthermore, this model can be embedded in a larger application. A good simulation engine combines efficiency, functionality and completeness to enable a model builder to construct a comprehensive, customized model of either a specific system or a class of systems. This paper presents CSIM18, a simulation engine which supports development of applications with efficient, embedded simulation models on a variety of system plat
ABSTRACT Energy-efficiency has become a critical parameter in the design of high-performance comp... more ABSTRACT Energy-efficiency has become a critical parameter in the design of high-performance computing systems. Typically, compute elements consume the most energy in current systems, with memories and interconnect networks close behind. This paper proposes an energy-efficient, high-connectivity, petascale computing system for the year 2020 timeframe by addressing the energy requirements of these three components. We start with projections based on purely evolutionary computer system design trends, then include the impact of breakthroughs in processor design, memory packaging and optical interconnect technologies. Based on these projections, we motivate the development for a 1-pJ/b optical intrasystem interconnect technology that significantly increases system interconnect bandwidth and relieves the distance-based energy dependence of electrical alternatives. We show that improvements in compute, memory, and IO, when simultaneously applied, become a vision for many-chip photonically-interconnected modules that could lead to an order of magnitude improvement in energy efficiency in the 2020 timeframe. The vision hinges on a high-density, energy-efficient optical link that can connect electronic compute and memory elements across short chip-to-chip distances while also capable of kilometer or longer spans across data centers. We discuss the power budget to enable such a link and review experimental progress toward creating an ultra-dense, hybrid-integrated low-power silicon photonic link that will enable this vision.
Proceedings of the fourth symposium on Operating system principles - SOSP '73, 1973
This paper presents a description and analysis of a task scheduling algorithm which is applicable... more This paper presents a description and analysis of a task scheduling algorithm which is applicable to third generation computer systems. The analysis is carried out using a model of a computer system having several identical task processors and a fixed amount of memory. The algorithm schedules tasks having different processor-time and memory requirements. The goal of the algorithm is to
WSC'99. 1999 Winter Simulation Conference Proceedings. 'Simulation - A Bridge to the Future' (Cat. No.99CH37038), 1999
ABSTRACT Developing a behavioral model of a system prior to implementing that system provides sig... more ABSTRACT Developing a behavioral model of a system prior to implementing that system provides significant benefits. In some cases, errors in the design can be detected and corrected prior to implementation. In other cases, having a means for predicting the performance of the system prior to its deployment may aid in the design, implementation and configuration of the system. This paper presents a case for developing simulation models of systems early in the design and implementation process. A number of examples illustrate the benefits which can accrue. 1 INTRODUCTION New systems are becoming increasingly more complex, offering more functionality, greater speed, and higher reliability, all at reduced costs. Along with these new requirements, there is also the fact that market windows are decreasing, time-to-market is decreasing, and the number of product cycles is increasing All of these factors mean that it is more important than ever that all facets of a newly designed system be validated a...
Proceeding of the 2001 Winter Simulation Conference (Cat. No.01CH37304), 2001
CSIM19 is the latest version of the system modeling toolkit from Mesquite Software. CSIM19 offers... more CSIM19 is the latest version of the system modeling toolkit from Mesquite Software. CSIM19 offers many features that enable a modeler to develop robust and realistic models of complex systems. These models represent the system as a collection of processes and resources; in most cases the processes mimic the behavior of the entities of the system as they compete for use of the system's resources. The paper presents CSIM19 and the new features in CSIM19. These features include new capabilities for managing processes at facilities, enhanced handling of messages at mailboxes, and performance improvements. Using the OptQuest optimization package with CSIM19 is also discussed. The article concludes with an extended example
Proceedings of the ACM annual conference on - ACM '72, 1972
This paper describes an experimental study of the performance of a large multi-programmed compute... more This paper describes an experimental study of the performance of a large multi-programmed computer system (the UT-l/CDC 6600 system at the University of Texas at Austin) under systematic variation of available resources and resource allocation algorithms. The experiments were carried out in a controlled and reproducible environment provided by a synthetic job stream generator. The experimental data was recorded by an event-driven software monitor which recorded a complete trace of system activities at the level of system defined events. The study relates resource utilization and queueing patterns to the metric of job completion rate. The experiments undertaken in these studies are single factor experiments. Compensatory reactions by this complex system to variation of individual resources are nonetheless revealed. The experiments also demonstrate the criticality of optimal scheduling of bottleneck resources and offer comparisons of the performance of multi-drive disk units under different conditions of availability and space assignment. The data gathering facility was also run on the production environment to determine base lines for comparison to the experiments as well as for an understanding of the production mode of operation of the system.
ABSTRACT In many cases, a simulation model of a system is used to evaluate alternative configurat... more ABSTRACT In many cases, a simulation model of a system is used to evaluate alternative configurations of that system, with the goal of finding the configuration which maximizes (or minimizes) the value of an objective while meeting all of the stated requirements. The CSIM18/OptQuest package automates this kind of search for the best configuration by combining a powerful simulation engine, CSIM18, and a state-of-the-art optimization package, OptQuest. This paper describes this integrated package for doing optimization and simulation. The paper concludes with two examples: finding the best configuration for a jobshop, and finding the best configuration for a web server.
Uploads
Papers by H. Schwetman