Massively parallel processing

The hardware situation continued to evolve. Personal computers became ever more powerful in terms of speed and the amount of random access memory (RAM) and hard drive capacity. The price of PCs continued to fall. Clusters of PCs were built. Use of the open-source Linux operating system spread in the 1990s. Distributed processing was developed so a long calculation could be farmed out to separate machines. Massively parallel processing was tried. All these changes meant that the days of the supercomputers were numbered. 

Self-assembly is a massively parallel process, and can normally involve very large numbers of components (a large crystallization might involve 1027molecules). Robotic pick-and-place methods for placement are limited by the fact that they are serial. Although they can be accelerated by using a number of robotic devices in parallel (for example, the multiple scanning probe heads of the IBM millipede 131), they cannot approach the number of molecules in a test tube, for example. 

The coupled code developed by Steefel and Lasaga (1994) for multicomponent reactive transport with kinetics of precipitation and dissolution of minerals has been developed further into the OS3D/GIMRT code (Steefel and Yabusaki, 1996). This model has been applied to reaction fronts in fracture-dominated flow systems (Steefel and Lichtner, 1998). Eurther developments for nonuniform velocity helds by Yabusaki et al. (1998) required the use of massively parallel processing computers, although ... the accuracy of the numerical formulation coupling the nonlinear processes becomes difficult to verify. ... 

M. Ishikawa, A. Morita, N. Teikayanagi, High speed vision system using massively parallel processing, in Proceedings of IEEE International Corference on Intelligent Robots and Systems, 373 (1992)... 

A number of papers have reported on CFD modeling with the commercial ANSYS/FLUENT software [13]. Several terms must be defined in conjunction with massive parallel processing prior to discussing different HPC applications. Cortex is a process that provides the user interface and graphics for FLUENT, that is, the graphic user interface (GUI) of ANSYS/FLUENT. ANSYS/FLUENT interacts with the host processor, which organizes and delegates the numerical tasks to different nodes. Each node possesses its own random-access memory (RAM), that is, 24 GB per node and 3 GB per core for the JuRoPA architecture. The host sends a request to the nodes, which store their results directly in parallel on the hard disk. 

Zone, O. Vanderstraeten, O. Keunings, R. A parallel direct solver for implicit finite element problems based on automatic domain decomposition. In Massively Parallel Processing Applications and Development, Dekker, L., Smit, W., Zuidervaart, J.C., Eds. Elsevier, Amsterdam, 1994 809-816. 

Driven by major advances in microfluidic biosystems, such as the development of micro total analysis systems, microfabricated cell sorters, microseparators for DNA and proteins, and cell-based assays, the microfabrication of biologically meaningful microenvironments is within the scope of recent activities. Bioassays that exploit miniaturized formats are intrinsically advantageous (in part because of their small sample volumes and massively parallel processing), but they heavily rely on defined sur-... 

The TNP architecture was inspired by the need for a low-cost massively parallel processing system that could emulate a large variety of neural models. The role envisaged for the TNP architecture is to serve as a tool allowing the exploration of neurocomputing problems. 

We shall see below how molecular computers, quantum computers, genetic algorithms, and cellular automata can allow for massive parallel processing of information, which overcomes the limitations of the standard, sequentially working von Neumann computers and conventional algorithms. 

---