China
Africa
Explore chapters and articles related to this topic
Force-System Resultants and Equilibrium
Published in Richard C. Dorf, The Engineering Handbook, 2018
format the data for raster display, and plot the data on a dot-matrix thermal line printer. No physical pen motion is required, so measurement speed is limited only by measurement electronics, paper speed, and printer resolution rather than pen motion. The more sophisticated display capabilities of the thermal printer allow incorporation of information other than raw display traces onto the strip chart. Grid lines may be printed at the same time as the traces, eliminating the alignment difficulty found with pen-based recorders, in which pen traces are aligned with grid lines preprinted on the chart paper. Other graphic and alphanumeric information may also be printed on the chart as it is created. For instance, the RMS Instruments GP300 (Figure 155.3) prints on a 300-mm wide strip at a resolution of 300 dots per inch
Acquiring and Processing Turbulent Flow Data
Published in Richard J. Goldstein, Fluid Mechanics Measurements, 2017
The central processing unit is the heart of the machine, where computation takes place. The data are organized as 8-, 16-, or 32-bit words. All computers have some fast random-access memory (RAM). The size of this memory typically ranges from 250 kbytes to 100 Mbytes in small laboratory computers. Computers also have read-only memory (ROM) for special-purpose memory, typically the “bootstrap” program for getting the system started and operating system or library programs. Computer systems usually have some mass storage like disks or tapes. Typical laboratory computers would have mass storage ranging from a floppy disk at 250 kbytes to hard disk storage of 40-80 Mbytes (or larger). Mass storage is generally slow in relation to computer cycle times; tapes are the slowest and least expensive, with access times of tens to hundreds of milliseconds, and hard disks are fastest, with access times in fractions of milliseconds. Once data have been located, data transfer is rapid, typically 10,000-100,000 bytes/s or faster. Tape storage is rapidly being displaced by disk storage, which continues to decrease in price, precipitously. Peripherals such as an alphanumeric keyboard, line printer, and plotter allow interaction with the experimenter. Access to the experiment is via digital or analog (through a D/A converter) signals sent to switches or actuators and via digital or digitized analog signals received from the sensors. The data bus provides an interface between the CPU and the peripherals and between the CPU and memory. Since the components on the data bus may be from different manufacturers, it is important that the bus (and interfacing) be standardized. Standard buses are IEEE-488, RS-232, S100, Q-bus, and Multibus. Most scientific equipment is designed for communication on the IEEE-488 or RS-232 interface bus. The data bus contains lines for transferring bits of data and bits of logic information needed to coordinate the data transfer.
Personal reflections on 50 years of scientific computing: 1967–2017
Published in International Journal of Parallel, Emergent and Distributed Systems, 2020
In 1969, after graduating from Evansville with majors in mathematics and psychology, I worked for another accounting firm (Wright, Griffin, and Davis, CPAs) in Ypsilanti, Michigan while a graduate student in mathematics at The University of Michigan. That work involved the niche IBM System 3, a machine with BCD (binary coded decimal) integer arithmetic, 32-bit words, a rotating disk, line printer, an assembler, and 96-column punched cards about 1/3 the size of IBM 80-column cards for I/O. I led a team (of mathematics graduate students, with no prior experience with computing) to design and build an interpreter called PROFIT (Periodic Reporting Of Financial Input Transactions) with functionality similar to ABLE, including a monitor and all the physical I/O device drivers. The input was on punched cards, the financial reports were printed, and the financial data was stored on a hard disk drive. The disk data structures involved binary trees, linked lists, and garbage collection, all created by students with no prior knowledge of any of these concepts. Years later, I was contacted about why PROFIT, rewritten in COBOL (which had become available for the System 3 by that time), might be running so much slower than the original assembly language version.