Khoa D. Huynh

Performance analysis of advanced I/O architectures for PC-based video servers

In the personal computing and workstation environments, more and more I/O adapters are becoming complete functional subsystems that are intelligent enough to handle I/O operations on their own without much intervention from the host processor. The IBM subsystem control block (SCB) architecture has been defined to enhance the potential of these intelligent adapters by defining services and conventions that deliver command information and data to and from the adapters. In recent years, a new storage architecture, the redundant array of independent disks (RAID), has been quickly gaining acceptance in the world of computing. In this paper, we discuss and present a performance analysis of the SCB architecture and disk array technology in typical video server environments. In particular, we would like to see whether a disk array can outperform a group of disks (of the same type, the same data capacity, and same cost) operating independently (not in parallel as in a disk array) in a video server environment where most disk I/O operations are large sequential reads.

A performance analysis of an object-based I/O architecture in a video server environment

In this paper, we present a performance analysis of how effective video server applications can be supported on personal computers (PCs) connected through a local area network (LAN). We considered both the standard 16-Mbit/s token ring and a 100-Mbit/s token ring, which follows closely the specifications for the Fiber Distributed Data Interface (FDDI). We examined three I/O architectures for a PC-based video server: an interrupt-driven I/O architecture, a peer-to-peer I/O architecture, and a concurrent, object-based I/O architecture that we proposed. The video server must support multiple MPEG-1 video streams at the same time to multiple clients on the LAN. We found that the network protocol layers require a lot of processing power, and that an implementation of our proposed I/O architecture, which takes advantage of the available power of the host processor to off-load I/O adapters, can deliver much better performance, and is more cost-effective, than other I/O architectures in a video server environment.

A performance analysis of the IBM subsystem control block architecture in a video conferencing environment

With the proliferation of new intelligent adapters, advanced bus architectures, and powerful microprocessors, the latest personal computers need a new architecture to coordinate the transfers of command and control information, data, and status between their system components in the most efficient manner. To this end, IBM has defined the Subsystem Control Block (SCB) architecture. In this paper, we would like to present our performance analysis of the SCB architecture in one of the most sophisticated multimedia applications today, the video conferencing environment. Our analysis demonstrates how effective the SCB architecture is in handling multiple video streams and file transfers concurrently in a typical video conferencing system based on personal computers.

Performance analysis of a peer-to-peer I/O architecture in video server environments

In the personal computing (PC) and workstation environments, more and more I/O adapters are becoming complete functional subsystems that are intelligent enough to handle I/O operations on their own without much intervention from the host processor. The IBM Subsystem Control Block (SCB) architecture has been defined to enhance the potential of these intelligent adapters by defining services and conventions that deliver command information and data to and from the adapters. The SCB architecture has two operating modes. The Locate Mode represents the conventional, interrupt-driven I/O protocol used in many current personal computers. The Move Mode embodies an advanced, peer-to-peer I/O protocol proposed for the next generation of personal computers. In this paper, we would like to discuss and present a performance analysis of the SCB architecture in typical video server environments. In particular, we study a video server capable of providing support for simultaneous MPEG-1 video streams to multiple clients on a 16-Mbps token-ring network. We also consider the performance impact of a hypothetical 100-Mbps token-ring technology on the video server performance.

A performance analysis of advanced I/O architectures for PC-based network file servers

In the personal computing and workstation environments, more and more I/O adapters are becoming complete functional subsystems that are intelligent enough to handle I/O operations on their own without much intervention from the host processor. The IBM Subsystem Control Block (SCB) architecture has been defined to enhance the potential of these intelligent adapters by defining services and conventions that deliver command information and data to and from the adapters. In recent years, a new storage architecture, the Redundant Array of Independent Disks (RAID), has been quickly gaining acceptance in the world of computing. In this paper, we would like to discuss critical system design issues that are important to the performance of a network file server. We then present a performance analysis of the SCB architecture and disk array technology in typical network file server environments based on personal computers (PCs). One of the key issues investigated in this paper is whether a disk array can outperform a group of disks (of same type, same data capacity, and same cost) operating independently, not in parallel as in a disk array.

A high-level performance analysis of the IBM subsystem control block (SCB) architecture

Abstract With the use of new intelligent adapters, advanced bus architectures, and powerful microprocessors, the latest personal computers need a new architecture to coordinate the transfers of command and control information, data, and status between their system components in the most efficient manner. To this end, IBM has defined the Subsystem Control Block (SCB) architecture. In this paper, we would like to present our modeling approach to analyze the operational and performance characteristics of this architecture and its two operating modes, the Locate Mode and the Move Mode. The models, which were implemented using the IBM Research Queueing Package (RESQ), are described, and the results obtained from the models are presented. The analysis of these results demonstrates the capability of the SCB architecture to handle large amounts of data in a LAN-server environment, and the superiority of the Move Mode over the Locate Mode as defined in the architecture.

A workload model for frame-based real-time applications on distributed systems

Abstract Many real-time control applications fall into the category of control cycle-driven systems. Usually called frame-based applications, the real-time aspect of this type of system is based on the concept of a control cycle, or frame, during which the input data is sampled, some critical processing on the data is performed, and the appropriate output is generated. If a frame is repeated at a sufficiently fast rate, the system is said to provide continuous control in real time. Several distributed processing approaches appear to be the most appropriate hardware architectures for meeting the requirements of these applications. To support the functional testing, performance analysis, and frame-based software development of these systems, a distributed real-time workload (DRTW) has been designed for use either as a parameterized workload for testing and performance evaluation purposes, or as a rapid software prototyping tool for frame-based applications on single- and multiple-node distributed systems.