Double-precision word

IlyperChem uses 16 bytes (two double-precision words) of storage for each electron repulsion integral. The first 8 bytes save thecom-pressed four indices and the second S bytes store the value of the integral. Each index lakes 16 bits. Thus the maximum number of basis fiinctions is 65,535. This should satisfy all users of IlyperChem for the foreseeable future. 

Because ol Lhe use of Lwo double-precision words for each in tegral. IlyperCbem needs, for example, ahoiil 44 MByles of computer mam memory and/or disk space Lo store the elecLroii repulsion inlejrrals for benzene wilh a double-zeta 6-i lG basis set. 

Because of the use of two double-precision words for each integral, HyperChem needs, for example, about 44 MBytes of computer main memory and/or disk space to store the electron repulsion integrals for benzene with a double-zeta 6-3IG basis set. 

IMAGE stores what is in KK(L) in KA. KK and KA have been defined by the program as double-precision variables. KK must be written with a Hollerith field specification using multiples of 8H. Care must be taken to leave trailing blanks to fill up the eight characters used in the double-precision word. 

Double-precision word A compu ter memory allocation used to store a single number that contains twice the number of bytes as a normal (single-precision) word. 

Double precision—Vdilue stored as two words, rather than one, representing a real number, but allowing for approximately double the number of significant digits. 

One way to increase precision, at the expense of computing speed, is to "chain" two or more words to provide a "double-precision" or "extended precision" word. The number of bits reserved to the exponent is sometimes increased, sometimes not, but the mantissa receives lots of extra bits from the extra chained words. This is done routinely for scientific work on 8-bit PDP-8 machines and on the early 16-bit PC s (IBM PC, PC AT, from Intel 80086 to 80186 to 80286 to 80486 CPUs), until 32-bit PC s became the norm (Pentium CPUs). 

In this calculation you will encounter problems caused by the finite word length of the computer. Usually you are shielded from such problems by the double precision of modern spreadsheets, but that is not enough for equations (5.8-29) and (5.8-30). 

The REAL 8 statement lists name arrays and words which are real double-precision variables. 

The arrays KA and KK have been defined as double precision by the program. The word used in KA in the calling sequence must either have been read into the program by a data statement, or written in the calling sequence in Hollerith format as an 8H field. 

The IBM System 360/65 for which this code is adapted uses a four—character word and the data names, which are multiples of six letters, are therefore double precision, using eight characters. An 8H format should be used and trailing blanks included. (In the case of the data statement, trailing blanks are inserted by the machine.)... 

The method used will depend on the situation. If a name is going to be used more than once it may be worth putting it in storage. Four words are automatically in storage and need not be specially entered they are END, H+, H20 and a blank. Note that when using the data statement it is necessary to specify NITRO as real double precision. This is because it would otherwise be integer, because it begins with an N. 

Digital Equipment Corporation (DEC) introduced a sophisticated minicomputer, the VAX 11/780, in the mid- to late-1970s. At prices of 300 k USD (and up), machines could be acquired that were as fast as 1960s multi-million dollar mainframes. Double precision meant 64-bit words, and multiple external disk drives could be added to provide hundreds of MB of storage. DEC was not the only producer of minicomputers, and before long, with different vendors and models, their use for computational chemistry became fairly widespread. 

From Fig. 9.16, we obtain c/33 1.05 kN, a value consistent with the value listed in the catalog (1.27 kN). The value might be somewhat lower than the true value because the bonding of the tube ends is not perfectly rigid. If one end of the tube is free, or both ends are free, the deformation pattern varies significantly at the end(s). The net end effect is to reduce the value of the double piezoelectric response. Even if the end-bonding condition is unknown, an accurate measurement of the temperature or time variation of the piezoelectric constant can still be achieved. In other words, if the piezoelectric scanner is calibrated by a direct mechanical measurement or by the scale of images at one temperature, then its variation can be precisely determined by the electrical measurements based on double piezoelectric responses. 

---