- AboutThis should describe the systems research collaboration, and present the overall research goals of the new group.
- PeopleHere are the different labs in the SRC…
- PublicationsA page where you will find categorized publications!
- ProjectsA page where you will find our projects
- ResourcesVarious resources for prospective students, current students, alumni. Maybe put something here about life in NYC and at Columbia…
Proceedings of the 5th International Workshop on Network and Operating System Support for Digital Audio and Video, April 1995
The advent of multimedia ushers forth a growing class of applications that must manipulate digital audio and video within well-defined timeliness requirements. Exist- ing processor schedulers are inadequate in supporting these requirements. They fail to allow the integration of these continuous media computations with conventional inter- active and batch activities. We have created a new scheduler that provides integrated processor scheduling for all classes of computational activities. Our solution achieves optimal performance when all timeliness requirements can be satisfied, and provides graceful degradation when the system is overloaded. Though unique in the degree to which it allows users control over the dynamic sharing of processing resources, the scheduler does not impose any draconian demands on the user to provide information he does not have or does not choose to specify.
Project Technical Report SML-94-0488, December 1994
Proceedings of the 4th International Workshop on Network and Operating System Support for Digital Audio and Video, November 1993
Applications that manipulate digital audio and video are rapidly being added to workstations. Such computations can often consume the resources of an entire machine. By incorporating a â€œrealtimeâ€ process scheduler, UNIX System V Release 4 (SVR4), the most common basis of workstation operating systems, claims to provide system support for multimedia applications. Our quantitative measurements of real application performance demonstrate that this process scheduler is largely ineffective and can even produce system lockup. While SVR4 UNIX provides many controls for changing scheduler performance, they are virtually impossible to use successfully. Furthermore, the existence of a realtime static priority process scheduler in no way allows a user to deal with these problems. This paper provides a quantitative analysis of real system behavior, demonstrates why it is not possible to obtain the kind of behavior desired with the mechanisms currently provided by the system, and presents modifications to improve the situation.
Proceedings of the 1992 Workshop on Volume Visualization, October 1992
Volume rendering is a useful visualization technique for under- standing the large amounts of data generated in a variety of scien- tific disciplines. Routine use of this technique is currently limited by its computational expense. We have designed a parallel volume rendering algorithm for MIMD architectures based on ray tracing and a novel task queue image partitioning technique. The combi- nation of ray tracing and MIMD architectures allows us to employ algorithmic optimizations such as hierarchical opacity enumera- tion, early ray termination, and adaptive image sampling. The use of task queue image partitioning makes these optimizations effi- cient in a parallel framework. We have implemented our algorithm on the Stanford DASH Multiprocessor, a scalable shared-memory MIMD machine. Its single address-space and coherent caches pro- vide programming ease and good performance for our algorithm. With only a few days of programming effort, we have obtained nearly linear speedups and near real-time frame update rates on a 48 processor machine. Since DASH is constructed from Silicon Graphics multiprocessors, our code runs on any Silicon Graphics workstation without modification.
Technical Report CSL-TR-92-537, August 1992
AI Technical Report 1139, September 1989
AI Memo 1165, September 1989
Technical Memorandum, December 1987
Memorandum for File, October 1986