The story of computers in the country’s east is deeply intertwined with the story of Prasanta Chandra Mahalanobis, who has been described as ‘a physicist by training, a statistician by instinct and a planner by conviction’.
In 1913, just as Mahalanobis was to set sail to India from England, Prof Macaulay, his tutor at King’s College, drew his attention to some bound volumes of Biometrika, a journal of statistics. Mahalanobis devoured the entire set of journals during his journey and discovered statistics as a new area of interest in his life. But, due to the First World War, Mahalanobis remained in India.
He joined Presidency College in Kolkata as a professor of physics. But his interest in statistics grew stronger, and he soon discovered interesting areas for application of statistics, such as meteorology and anthropology. In the 1920s, he ran a workshop called ‘statistical laboratory’ from his university room and conceived the idea of a statistical institute in the country.
The Indian Statistical Institute was thus founded in 1931 and soon developed a new statistical technique for estimating acreage and yield of crops in a large region by random sampling. This was applied to estimate the jute crop in the province of Bengal in 1937.
In 1950, Mahalanobis helped establish the National Sample Survey (NSS) for collection of socio-economic data such as consumer expenditure, public opinion, forecasts of acreage and yield of crops. These surveys provided the copious amount of demographic and economic data required for running the Mahalanobis model of the second Five-Year Plan.
Mahalanobis reasoned that the only way to keep up with so much data was by means of a computer. He said, ‘We must proceed with electronic computers with all possible speed. Otherwise, we will never be able to cope with the tremendous volume of primary information which is accumulating... I know that real planning would require the use of such computers.’
Early Computers at ISI
The story of ‘computers’ at ISI actually began in the 1930s when the jute crop surveys were undertaken. These were unusual computers – unusual from the point of view of the conventional understanding of what a computer is. They were actually humans. Mahalanobis designated a few statistical staff members as computers, who calculated and tabulated the large-scale sample surveys conducted at the institute.
Mahalanobis also realised the importance of scientific instruments to aid the statistical efforts of the institute. In 1953, the first general report on the NSS, conducted under Mahalanobis’s leadership, noted: ‘Since much of the work of tabulation and analysis of the primary data was to be done by tabulating machines, training was also given to a large number of punchers and verifiers in the Institute... Arrangements were made to hire the latest types of tabulating machines from the International Business Machine Corporation (IBM) of New York; and by the latter part of 1951, the Institute had two new models of IBM tabulators, a new multiplier (an electronic calculator) and several sorters (for sorting punched cards), reproducers (for copying information from one deck of cards to another), etc , in addition to some of the machines of the British Tabulating Machine Co which the institute had been using for some time. An Electronic Statistical Machine (a high-power combined sorter-tabulator) was also rented from the IBM.’
That same year, the Electronic Computer Division at ISI built, with salvaged materials from the disposal depots of Kolkata’s Chandni Chowk market, India’s first electrical analogue computing machine that could solve simple linear equations. Around the same time, V Rajaraman, a student at the Indian Institute of Science, was working on a differential analyser.
According to the ISI annual report for 1952–53,12 these machines were used for a variety of surveys, such as the National Sample Surveys, model sampling experiments in connection with crop surveys (jute, cinchona, etc.) and population surveys (of Mysore, refugees in West Bengal, etc.).
The HEC-2M Comes to ISI
As we saw earlier, Mahalanobis soon realised the importance of getting a digital computer with enhanced capabilities to process data that India was collecting as part of its planning process. In 1954, he convinced the Soviet Ambassador Menshikov to help India acquire a digital computer. Eight months later, the USSR and the United Nations agreed to fund India’s purchase of a Hollerith Electronic Digital Computer-2M (HEC-2M) from the British Tabulating Machine Company for a princely amount of Rs 2 lakh.
Describing the events surrounding this digital computer, Rajaraman said: ‘In 1955, ISI got a digital computer imported from England, made by the British Tabulating Machine (BTM) Company. Two of their professors, Amresh Roy and Monimohan Mookerji, went to England and worked with Prof Booth, who was building that machine at Birkbeck College London. BTM was not willing to instal and maintain the machine in India, because only one machine was coming to the country. The ISI professors came back to India after their training, and sometime in the middle of 1955 when the machine arrived, they installed it and made it work. The machine had all of a 1024-word memory. It was a drum memory. (It did not have a printer or a tape and used punched cards.) The arithmetic was elementary and there was no language; they had to use machine language. They solved some interesting problems for ISI. That’s probably the first machine that came to India.’
Indeed, the HEC-2M was the first digital computer not only in India but also possibly in mainland Asia.
In 1958, with a grant from the United Nations Technical Assistance Administration (UNTAA), ISI got its second digital computer, the URAL. Unlike the HEC-2M, the URAL came with a complete set of detailed manuals and eight Russian engineers to help with the installation and training of personnel. The URAL was a 32-bit-word-sized 2Kb-memory machine with a horizontal magnetic tape, punched celluloid tape and a printer. Soon the computer division had around thirty people working full time to crunch NSS data and create optimal economic plans for the Planning Division.
In 1961, ISI, in collaboration with Jadavpur University, decided to design and build two solid-state electronic digital computers. The machines were planned to be character-based, having a syllable structure, unlike the word structure of the HEC-2M and the URAL, and was expected to be adaptable to automatic programming using a universal programming language such as ‘ALGOL’.
The computer was commissioned in 1965-66 and was christened as ISIJU-I (ISI Jadavpur University - I). However, the project was not highly successful – it was not able to get the card reader and line printer free of charge from the United Nations, and the input-output equipment that was designed consequently did not function satisfactorily. The computer’s projected data link could not be operated continuously due to serious fluctuation in the voltage of electric current at the institute.
Soon, ISI procured an IBM 1401, and the computation services at ISI consisted mostly of programming on this computer. The research and training school at ISI imparted training in efficient programming and the operation of tabulating and computing on machines. Their three-month evening course on ‘Punched Card Systems’ in Kolkata became popular. ISI was a strong contender in the race to become the first National Computational Centre in India. But it lost that race to TIFR.
By the 1960s, the east had ceded its leading position to the west of India. Institutions like TIFR and AEE, and people like Homi Bhabha, took centre stage.
Excerpted with permission from Against All Odds: The IT Story Of India’, Kris Gopalakrishnan, N Dayasindhu, and Krishnan Narayanan, Penguin.