166 Part 2 The instruction-set processor: main-line computers

concerning them from the engineering point of view are appropriate.

The Census computer was given two tests; the first, a test of its computational ability; the second, a test of its input-output system which particularly stressed the tape reading and recording abilities.

The Central Computer Acceptance Test A consisted of two parts. During Part 1, every available internal operation, except input-output operations, was performed. Among these operations were addition, subtraction, comparisons, division, and three different types of multiplication operations. Each of the arithmetic operations handled a pair of 11-decimal digit quantities. Altogether there were about 2,500 operations in the routine, yet the entire routine required only 1.26 seconds to do. The routine was performed 808 times in 17 minutes making a total of about 2,000,000 operations in all.

The second part of Test A included the solution of a heat distribution equation, a short routine involving the input-output device and a sorting routine. The sorting routine arranged ten numerical quantities each containing 12 decimal digits in correct numerical order in about 0.2 second. All three routines took a total of 11/2 minutes to perform. They were performed twice for each test and when added to Part 1 made a total of 20 minutes for unit test A.

The Acceptance Test B examined the input-output tape devices (Uniservos). During the first part of Test B, 2,000 blocks or about 1.4 million digits, which included every available character (numeric and alphabetic) were recorded on a tape and then read back into the computer with the tape moving backward. The information read back was then compared with the original data read out. The recording operation required about 4 minutes while reading back and comparison required about 8 minutes. The second part of Test B consisted of recording and reading over one spot of tape for 700 passes in order to determine the readability of tape as it wears. This test required 13 minutes and when combined with Part 1, made a total of approximately 25 minutes for Test B. This test was repeated 19 times.

The first test run passed in 6.6 hours (minimum theoretical time: 6.0 hours) and the second test was passed in 9.47 hours (minimum theoretical time: 7.45 hours). Of the 2.02 hours down time, 1.45 hours were accumulated at one time with the remaining 058 hours spread over the rest of the test.

The Uniprinter test required that a block of information (60 words) be printed 200 times in tabular form. The minimum time for printing was five hours. The test was passed in 6.16 hours.

The card-to-tape test required that ten good reels of tape be produced in 12 hours. There were certain restrictions as to reading accuracy and other criteria of reproducing ability which defined "good" reels. In 10 hours, the converter had prepared over 15 reels, 14 reels had been tested, 11 of the 14 were found satisfactory and the converter was accepted for payment.

Although the test was run on only one of two converters, the Bureau of Census put both card-to-tape machines into operation and after six months of use, the acceptance test was run on the second card-to-tape converter. This test differed to some extent from the first test in that the Census Bureau was satisfied with the reading ability of the machines and did not require a digit-by-digit verification of the information. However, a new stipulation was added that, after the engineers had checked the converter out preparatory to running the test, the converter was to be used in actual operation for eight hours before doing the remainder of the test with no engineering intervention between the two portions of the test. The first part was run on Friday, October 5, 1951; the device remained idle Saturday and Sunday and was turned on Monday morning to complete the test. It passed with flying colors, preparing ten acceptable reels (out of ten reels) plus two decks of check cards in slightly less than 7 hours. Both card-to-tape converters now are in Washington and the remainder of the system is in operation by the Bureau of the Census on the Eckert-Mauchly premises in Philadelphia.

The first UNIVAC system now has been operating for approximately 8 months. In that time, much has been learned about how UNIVACs should be operated and maintained. The situation has been somewhat complicated by having to shake down the equipment while in the customer's possession; that is, there were certain faults in the system from both engineering and production standpoints which could only become apparent in the course of time and under actual operation conditions. For example, weak tubes or faulty solder joints did not reveal their presence at the time of installation. Another type of difficulty only became apparent under certain duty cycle conditions imposed by various types of problems. Because only certain problems present this particular duty cycle, these troubles remained in the machine causing intermittent stoppages until they could be tracked down.

Patient isolation and elimination of such problems, most of which have occurred only with conditions of operation infrequently encountered, is a powerful, though sometimes painful proving ground for the engineering group charged with such responsibility. The experience and depth of judgment acquired by such a group in the course of performing such work have become unmistakably apparent in the already noted improved performance of following UNIVACs and generally advanced ability to predict