Master-slave model

We show in Figs. 2 and 3 the speed up for a Davidson iteration obtained with the PVM [17] ( Parallel Virtual Machine ) and PVMe [18] ( PVM enhanced ) message passing interfaces respectively. Speed-up factors are here relative to the sequential version of the program (n=l), and the theoretical mciximum has been defined according to the expression s n) — n — I appropriate to the used master/slave model. 

Fig. 3.57. Client-Server and the Master-Slave model for PC cluster applications...

Two different strategies may be employed to partition the work among processors. The first approach, dynamic load balancing, often uses a master-slave model where one processor, the master, keeps a list of tasks to be executed and apportions these tasks to the other processors, the slaves (the master may be a slave as well). When a processor has completed a task, it reports back to the master and requests a new task. This dynamic task allocation requires interprocessor communication, but has the advantage that load balance may be obtained even for problems where the execution time for individual tasks is variable and not known in advance. 

A totally different approach respects the idea that a Virtual Reality application that has basically the same state of its domain objects will render the same scene, respectively. It is therefore sufficient to distribute the state of the domain objects to render the same scene. In a multi-screen environment, the camera on the virtual scene has to be adapted to the layout of your projection system. This is a very common approach and is followed more or less, e.g., by approaches such as ViSTA or NetJuggler [978]. It is called the master-slave, or mirrored application paradigm, as all slave nodes run the same application and all user input is distributed from the master node to the slave nodes. All input events are replayed in the slave nodes and as a consequence, for deterministic environments, the state of the domain objects is sjmchronized on all slave nodes which results in the same state for the visualization. The master machine, just like the client machine in the client-server approach, does all the user input dispatching, but as a contrast to the client-server model, a master machine can be part of the rendering environment. This is a consequence from the fact that all nodes in this setup merely must provide the same graphical and computational resources, as all calculate the application state in parallel. 

Several other mechanisms such as the master-slave hypothesis (Callan and Lloyd, I960 Callan, 1967) or the gene expansion-contraction model (Brown et ah, 1972) have been proposed to account for the apparent identity in base sequence of serially repeated genes, e.g., the rRNA genes. However, a discussion of these and other rectification mechanisms would go beyond the scope of this review. 

Typically, parallel quantum chemistry algorithms require that several pairs of processors communicate simultaneously. Potentially, two messages will vie for the same communication path. This will lead to additional efficiency loss not accounted for in our simple model. While difficult to model, some simple rules can be checked to make sure that there are no obvious communication channel contention problems. First, the amount of communication handled by a single processor is limited by each communication channel s bandwidth, /p, and the number of channels possessed by each processor. This is particularly relevant to master-slave schemes where one processor is assigned more communication than the others. Second, the aggregate bandwidth of the network cannot be exceeded. This is particularly pertinent in networks with very limited connectivity, such as a bus network. 

The VMC and DMC methods are implemented in the casino code [44] (version 3.0), with which we intend to carry out this project. CASINO is written in Fortran 95 and parallelised using MPI with a master/slave program model, casino has been in existence for more than thirteen years and has been used on a wide variety of high-performance computer platforms. There are 360 registered users of the code. CASINO requires only the MPI library. 

The development of a calibration model is a time consuming process. Not only have the samples to be prepared and measured, but the modelling itself, including data pre-processing, outlier detection, estimation and validation, is not an automated procedure. Once the model is there, changes may occur in the instrumentation or other conditions (temperature, humidity) that require recalibration. Another situation is where a model has been set up for one instrument in a central location and one would like to distribute this model to other instruments within the organization without having to repeat the entire calibration process for all these individual instruments. One wonders whether it is possible to translate the model from one instrument (old or parent or master. A) to the others (new or children or slaves, B). 

This method can be considered a calibration transfer method that involves a simple instrument-specific postprocessing of the calibration model outputs [108,113]. It requires the analysis of a subset of the calibration standards on the master and all of the slave instmments. A multivariate calibration model built using the data from the complete calibration set obtained from the master instrument is then applied to the data of the subset of samples obtained on the slave instruments. Optimal multiplicative and offset adjustments for each instrument are then calculated using linear regression of the predicted y values obtained from the slave instrument spectra versus the known y values. 

Standardizing the coefficients of the model entails modifying the calibration equation. This procedure is applicable when the original equipment is replaced (situation 1 above). Forina et al. developed a two-step calibration procedure by which a calibration model is constructed for the master (F-X), its spectral response correlated with that of the slave X-X) and, finally, a global model correlating variable Y with both X and X is obtained. The process is optimized in terms of SEP and SEC for both instruments as it allows the number of PLS factors used to be changed. Smith et al. propose a very simple procedure to match two different spectral responses. 

Standardizing the spectral response is mathematically more complex than standardizing the calibration models but provides better results as it allows slight spectral differences - the most common between very similar instruments - to be corrected via simple calculations. More marked differences can be accommodated with more complex and specific algorithms. This approach compares spectra recorded on different instruments, which are used to derive a mathematical equation, allowing their spectral response to be mutually correlated. The equation is then used to correct the new spectra recorded on the slave, which are thus made more similar to those obtained with the master. The simplest methods used in this context are of the univariate type, which correlate each wavelength in two spectra in a direct, simple manner. These methods, however, are only effective with very simple spectral differences. On the other hand, multivariate methods allow the construction of matrices correlating bodies of spectra recorded on different instruments for the above-described purpose. The most frequent choice in this context is piecewise direct standardization... 

Standardizing the predicted values is a simple, useful choice that ensures smooth calibration transfer in situations (a) and (b) above. The procedure involves predicting samples for which spectra have been recorded on the slave using the calibration model constructed for the master. The predicted values, which may be subject to gross errors, are usually highly correlated with the reference values. The ensuing mathematical relation, which is almost always linear, is used to correct the values subsequently obtained with the slave. 

This method is probably the simplest of the software-based standardization approaches.73,74 It is applied to each X-variable separately, and requires the analysis of a calibration set of samples on both master and slave instruments. A multivariate calibration model is built using the spectra obtained from the master instrument, and then this model is applied to the spectra of the same samples obtained from the slave instrument. Then, a linear regression of the predicted Y-values obtained from the slave instrument spectra and the known Y-values is performed, and the parameters obtained from this linear regression fit are used to calculate slope and intercept correction factors. In this... 

The first five and the seventh were suggested first by Walker (1968) and Walker et al. (1969), the sixth by Britten and Kohne (1968) and the eighth by Britten and Davidson (1969) and Georgiev (1969). The ninth is implicit in models of multistranded chromosomes or chromosome structure involving tandem duplication of genes of the master and slaves type (Callan, 1967, Whitehouse, 1967 Thomas, 1971 Edstrom, 1968 Beermann, 1967). 

The model - nature itself - uses parallel structures in generational parent-to-child succession. As a result, it makes sense to apply this parallelity to the structure in the optimization program system as well. The EVOBOX code works on the processor as a so-called master and is responsible for mutations and selections, while other processors configured as slaves have to compute the quality functions of the different individuals. 

The dynamic behaviour of batch process units changes with time and this makes their precision control difficult. The aim of this paper is to highlight that the slave process of batch process units can have a more complex dynamics than the master loop has, and very often this could be the reason for the non-satisfying control performance. Since the slave process is determined by the mechanical construction of the unit, the above mentioned problem can be effectively handled by a model-based controller designed using an appropriate nonlinear tendency model. The paper presents the structure of the tendency model of typical slave processes and presents a case study where real-time control results show that the proposed methodology gives superior control performance over the widely applied cascade PID control scheme. 

---