SIMULATION

For shared bus multiprocessors operated in multiprogramming mode, the addition of second-level caches tends to significantly increase system performance. Trace-driven simulation was employed to obtain performance measurements over a range of system parameters, with the cache sizes at both levels being the parameters of most interest. For both light and heavy system loading, the addition of second-level caches was found to boost system performance. For heavily loaded multiprocessor cases, the workload-averaged percentage increases in performance ranged from 187% with 32k byte first-level caches to 507% with 4k byte first-level caches when 128k byte second- level caches were added. The main memory configuration and number of processors largely dictates the performance of a shared bus multiprocessor running in multiprogramming mode. The addition of larger second-level caches to the system results in increased system performance over a range of system configurations and workloads.

Simulation techniques are discussed to build test patterns to diagnose hardware faults in digital circuits. Only combinational circuits without redundant wires or undetectable faults are considered. All faults are assumed to be either stuck-at type faults or bridge-faults. An algorithm is presented to build test patterns, based on the circuit structure. The set of tests obtained from this algorithm is shown to have the diagnostic resolution to diagnose all the stuck-at type multiple faults and bridge-faults in the circuit. The simulation program accepts the circuit description as an input. It will store the details of the circuit in data structures that will enable the program to traverse both to the primary inputs and to the primary outputs from any wire. It is illustrated how the simulation program builds the test patterns, finds the fault ranges of each test and finally builds the fault dictionary of the circuit. A theory is presented to prove the completeness of the diagnostic resolution of the obtained test-set.

Clinical laboratories must balance staff and equipment utilization with specimen through put and turnaround time while reducing errors, costs and employee health hazards. Simulation can help to anticipate the effect of new technol ogy, alternative operating procedures and lab capacity changes.

GPSS/H-386 was used to simulate the analyzer area of a hospital clinical chemistry lab and the time-consuming pre-analytic processes of both a large hospital lab and a commercial laboratory outpatient facility. The processes modelled were the arrival of patients and receipt of specimens into the lab, data entry, blood-drawing, centrifugation and aliquotting. Thegreatest challenge was accurate modelling of a system driven by human decision making with flexible task priorities.

Experiments in the outpatient facility indicated that staff resources could be pooled instead of having specific job assignments without a significant effect on performance measures, and that an amalgamation of two outpatient labs could reduce staffing. These conclusions were not intuitively obvious to the lab managers.

Modern signal processing methods strive to maximize signal to noise ratios, even in the presence of severe noise. Frequently, real world data is degraded by under sampling of intrinsic periodicities, or by sampling with unevenly spaced intervals. This results in dropout or missing data, and such data sets are particularly difficul to process using conventional methods. In many cases, one must still extract as much information as possible from a given data set, although the available discrete data is sparse or very noisy. In such cases, we have found the algorithms derived from Chaos and fractal theory to represent a viable alternative to traditional spectral analysis. The data analysis techniques discussed in this work include phase space reconstruction, Poincare projections radius of gyration exponent, artificial insymmetration patterns (AIP), Liapunov spectra, correlation techniques, R/S analysis, K-factor, fractal statistics, maximum entropy method, and wavelets.

This paper describes the manner in which a windows-oriented, graphical user interface can facilitate the fitting of statistical distribution functions to data representing random events. A multi featured program that can display data in the form of histo grams, fit specified distribution functions to the data and determine a distribution function that best fits the data is described and illustrated with graphical examples. The results are displayed both graphically and in tabular form. Additional features are also described. The program has been written to be compatible with the SIMAN graphical user interface, though the results are general and can be applied within any particular graphical environment.