Computers, India: history, milestones
This is a collection of newspaper articles selected for the excellence of their content.
India’s first computer
India’s first computer
The first Indians to undergo training on computers were probably professors Amresh Roy and Monimohan Mookerji of the Indian Statistical Institute (ISI) in Kolkata in the mid-1950s. ISI purchased a HEC2M made by British Tabulating Machines (BTM) and designed by Andrew Booth at Birkbeck College London. The Indian professors trained with Booth on how to program and maintain this computer. BTM felt it would sell only one computer in India, and it was not economically viable to station their service team in India.
In the 1960s, many young scientists recruited for India’s nuclear energy and space programmes were trained on the indigenous TIFRAC computer built by the Tata Institute of Fundamental Research (TIFR). Abdul Kalam, who became the President of India, recalled that one of his first assignments when he joined the space programme was to work with the TIFRAC team. TIFR’s and Bhabha Atomic Research Centre’s capability in computer hardware were instrumental in setting up a public sector company, Electronics Corporation of India Ltd (ECIL). The TDC 12 from ECIL was India’s commercially produced computer.
The first programmes to build Indian capability in high-level programming probably began in the early 1960s in IIT Kanpur. Since they were set up with US support, faculty from leading US universities visited IIT Kanpur to set up labs and teach. An IBM 1620 computer arrived in IIT Kanpur in 1963 and was among the first computers in India that could be programmed with a highlevel language like Fortran. The visiting American faculty along with the now legendary Indian fac- ulty, V Rajaraman and H N Mahabala, set up the IIT Kanpur computer lab and devised a ten-day course. The course covered Fortran programming, numerical methods, and introduction to computer architecture. This was open to professionals in government and industry. F C Kohli, the first CEO of TCS, attended this course.
The first course in programming for undergraduate students was the TA (Technical Arts)-306 in IIT Kanpur. Abhay Bhushan, an IIT Kanpur student of this era, went on to write the original specification for the seminal File Transfer Protocol while at MIT. In the mid-1960s, IIT Kanpur introduced the MTech programme in electrical engineering with a specialisation in computer science. N R Narayana Murthy, the founder of Infosys, is a graduate of this programme.
With the growing interest in programming in India, Rajaraman’s lecture notes were published as the book ₹Principles of Programming’ in the late 1960s. The publisher printed it on newsprint since the author insisted that the book should be affordable to students. IIT Kanpur went on to start one of India’s first undergraduate BTech courses in computer science in 1978.
In 1978, the department of electronics set up a manpower development committee headed by Rajaraman that recommended a new academic programme called the Master of Computer Applications (MCA), as a post BSc qualification. This improved the job prospects for young Indians with a BSc degree and provided a boost to the supply of trained IT professionals. A famous graduate of the MCA programme is N Chandrasekaran, who was the CEO of TCS and is now the chairperson of Tata Sons.
Mahabala moved to IIT Madras in the early 1970s to set up the computer science department there. IIT Madras got an IBM 370, which was one of Asia’s most powerful computers in that era and started an MTech programme in computer science. IIT Madras encouraged students with an MSc degree to enrol. This provided an opportunity for Kris Gopalakrishnan, co-founder of Infosys, to pursue this MTech programme.
N Dayasindhu is co-founder of itihaasa Research and Digital.
In-company training programmes evolved from the early 1970s when TCS invited the faculty of IIT Kanpur to teach computer programming to their college recruits. Later, they invited H N Mahabala to set up a training programme on programming and systems design and this became the foundation for the industrial scale in-house training programmes for fresh college recruits in TCS. A pinnacle of Indian IT’s capability building is Infosys’ Global Education Centre in Mysuru. It rivals the campuses of the best universities in the world in terms of its state-of-the-art learning and recreational facilities, and can simultaneously train about 13,500 professionals.
HCL and its precursors
For years, the Indian computer industry has had arms swinging in both directions, hardware and software. The big boys of the Indian computer scene have been adept at both. It may come as a surprise to many that some of the leading software companies of today actually started out as hardware firms or had a hardware connect. The birth of Infosys, for example, can be traced to Mumbai-based Patni Computers, which used to be a distribution company for Data General (one of the first minicomputer firms of the late 1960s in the US).
Founder Narendra Patni became one of the early software exporters when he started getting business for software. In 1977, Narendra Patni appointed NR Narayana Murthy to head the software division. Murthy left Patni Computers in 1980 along with six colleagues to found Infosys. Just like Murthy, other Infosys co-founders, including Nandan Nilekani and Kris Gopalakrishnan, too have a hardware connect, having started their career at Patni Computers. Hinditron, founded in 1966 and now into software, telecom and more, too started with selling digital hardware and PCs. Harish S Mehta, co-founder of Nasscom, was in hardware initially. Tata Burroughs was formed following a collaboration between the Tata Group and Burroughs in 1977. While Burroughs was among the largest computer manufacturers in the world, its joint venture with TCS opened for Tatas many global opportunities including TCS’s first big IT outsourcing contract – to upgrade a hospital accounting system for about $24,000.
HCL too has its origins embedded in hardware. We had set up a hardware manufacturing plant in Silicon Valley, US. But due to a US government specification, that hardware could not be supplied, and a large loan from ICICI Bank became a question mark. The product that we wanted to sell in America was a multiprocessor – a Unix-based minicomputer. The OS we created for this multi-processing hardware was absolutely unique – a fine grained Unix, the first in the world. Since we could not continue with the hardware business in America, we pivoted towards software, strengthening our R&D base. Our Unix capability led us to the software business. Wipro too moved to software because of its hardware capability. That’s how our homegrown firms started competing with biggies like Accenture and Capgemini.
Today, all the top names in theindustry carry out engineering and R&D exports, where the knowledge of hardware and embedded software is used to create products for global markets. There is also development of semiconductor chips. The kind of skills we created opened the paths for software businesses. At HCL, we started our Singapore subsidiary in 1980 to manufacture computers. The market was like in India, but also wanted solutions. This meant providing also the application software, such as for finance and accounting, sales order processing, inventory control. So, to support the hardware business, we started a software export division in Madras (now Chennai) in 1980. We were amongthe earliest to do software in India for global companies.
Back in the day, we would carry out a system study of the requirements of our Singapore customers and send back the system report to our software export division in Madras, that created specialised tools. Those were very early days in the automation of the whole process of programming, to export the software faster. Today, there’s a lot of talk about automating software, but we did it in 1980, when communication links left much to be desired. We would make the software and courier the floppy disks, which would often get corrupted due to Customs’ x-rays. The process would painstakingly be repeated to get it sent to the customer.
A lot of capability got built and honed over time as a result of designing and making our own software and hardware. We created our own word processing software, our own relational database, much before Oracle did! We realised early on that building our own skill capital would lead us to long-term sustainability.
Given the exploding demand for electronics now, all these skills will help us in becoming the product capital of the world, backed by favourable geo-politics. Ajai Chowdhry is co-founder of HCL, chairman of EPIC Foundation, and author of Just Aspire
Supercomputers and India
PARAM 8000: India’s first Supercomputer
A home-made, super-duper supercomputer
The PARAM supercomputer began its life as a result of a technology embargo that had been placed on India. As the story goes, in 1987, during a high-technology meet between India and the US, the then US president Ronald Reagan refused to sell to then prime minister Rajiv Gandhi the latest CRAY supercomputers being developed in America. Instead, the Indian premier was offered an outdated machine, and also warned against its use for any purposes other than weather forecasting. This was not without good reason-supercomputers are an essential step toward modern weapons systems. For example, the calculations required to construct an intercontinental ballistic missile cannot be performed on an ordinary computer; it would take far too long. In 1988, for the purposes of developing self-sufficiency in such matters, India established the Centre for Development of Advanced Computing, or C-DAC, with Dr Vijay Bhatkar as director. With a deadline of three years, and a budget of around Rs 30 crore, PARAM 8000 was born in 1991.
Pune supercomputers among world’s fastest
Swati Shinde Gole TNN 2013/06/20
Pune: Two computers from Pune feature among the top 100 in the list of 500 fastest supercomputers in the world.
While the Indian Institute of Tropical Meteorology (IITM) supercomputer stands 36th in the list, Param Yuva II, developed by the Centre for Development of Advanced Computing (C-DAC), has bagged the 69th position. The IITM supercomputer is yet to be installed while the Param Yuva II became operational this February.
The list of the world’s top 500 supercomputers was announced on Tuesday at the launch of the opening session of the International Supercomputing conference in Leipzig, Germany.
The Param Yuva II has a capacity of 524 teraflops and within three weeks of the launch, it was already running at 70% of its capacity. Precise weather forecasting, faster tapping of natural resources in the sea and designing of customized drugs are possible using Param Yuva II. The IITM supercomputer will start functioning in the next two to three months.
2016: India ranked 9th in global race for supercomputers
What is a supercomputer?
Supercomputers are broadly defined as the fastest computing systems at any given time and are used for scientific purposes that require handling of troves of data at high speed. This includes testing mathematical models using thousands of variables to infer complex phenomena like weather, climate change, nuclear reactions, origin of the universe and so on.
How is the computing power of a supercomputer measured?
The standard unit to measure computational power is FLOPSfloating point operations per second. It is the number of mathematical operations involving fractions that a computer can do per second. For the most basic computers and smartphones, the computational ability is a few megaflops (more than a million operations per second). Chinese supercomputer Sunway TaihuLight -which is ranked the fastest in the world by Top500 (a supercomputer ranking project) -has a computational ability of 93 petaflops, more than a billion times faster than normal computers. Recently it was announced that Japan will spend $173 billion to bu ild the world's fastest supercomputer system with a capacity of 130 petaflops.
What problems were faced by supercomputer pioneers who worked to enhance computer speed?
First, the maximum speed at which electronic signals can travel cannot be faster than the speed of light. Second, because of frequent receiving and transmission of these signals, it was important to have a suitable cooling technique to control the temperature of the system. Both barriers were tackled by decreasing the lengths these signals were re quired to traverse by using circuit boards and innovation in cooling techniques.Other advancements involved introduction of vector arithmetic in computing.
When was the world's first supercomputer built?
In 1965, CDC 6600, a supercomputer designed by US engineer Seymour Cray was installed at the CERN in Geneva, Switzerland. CDC 6600 had a computational capaci ty of three megaflops. In the 1970s, the series CDC 7600 was introduced, which was10 times faster than the earlier version. In 1976, the Cray-1supercomputer was installed in the Los Alamos laboratory with a computation capacity of 160 megaflops.
What were the limitations of Cray's designs?
Cray's designs used expensive technology and liquid immersion technology to achieve high speeds. W Daniel Hill, an MIT graduate proposed a decentralised control instead of one CPU, which meant several inexpensive processors could be arran ged to achieve speeds comparable to the most expensive supercomputer designed by Cray. Hill designed the CM-1 in 1985, which used thousands of inexpensive processors to achieve the same speed as Cray's computer.
Which countries are the world leaders in supercomputers today?
Of the 500 fastest supercomputers, the highest numbers are in the US and China, each having 171 such systems. Germany , Japan and France each have 20 or more such systems. India is ranked 9th with five of its supercomputers listed in Top500.
2019/ 2 Indian computers among the world's 100 fastest
2 Indian super computers in the world's 100 fastest list
The list of world's fastest super computers is out. The bi-annual list called TOP500 is dominated by China and the US. There are also two Indian supercomputers in the list, ranked at No. 57 and No. 100. Here's over to India and the world's fastest supercomputers …
Summit, installed at Oak Ridge National Laboratory United States
The world's fastest super computer is IBM-built. It boasts of an HPL result of 148.6 petaflops.
Sierra, installed at at Lawrence Livermore National Laboratory in the United States
Ranked at No. 2, Sierra is another IBM-built supercomputer in the Top 10. The second-ranked Sierra system comes in at 94.6 petaflops.
Sunway TaihuLight, is installed at the National Supercomputing Center in Wuxi, China
At no. 3 is Sunway TaihuLight supercomputer, with an HPL mark of 93.0 petaflops. TaihuLight was developed by China’s National Research Center of Parallel Computer Engineering & Technology (NRCPC).
Tianhe-2A (Milky Way-2A), installed at the National Supercomputer Center in Guangzhou, China
The fourth fastest super computer in the world is Tianhe-2A (Milky Way-2A), a system developed by China’s National University of Defense Technology (NUDT). It boasts of GPL result of 61.4 petaflops.
Frontera, installed at the Texas Advanced Computing Center of the University of Texas in the United States
At No. 5 is Dell C6420 super computer Frontera. It has an HPL result of 23.5 petaflops. Powered solely by Xeon Platinum processors.
Piz Daint, installed at the Swiss National Supercomputing Centre in Lugano, Switzerland
At No. 6 is Piz Daint, a Cray XC50 supercomputer. At 21.2 petaflops, it remains the most powerful system in Europe.
Trinity, installed at Los Alamos National Laboratory and Sandia National Laboratories in the United States
Trinity, a Cray XC40 system, is ranked number seven, with an HPL performance of 20.2 petaflops.
ABCI, installed at Japan's National Institute of Advanced Industrial Science and Technology
Listed at No. 8, AI Bridging Cloud Infrastructure (ABCI) is Fujitsu-built system with an HPL result of 19.9 petaflops.
SuperMUC-NG, installed at Leibniz Supercomputing Centre in Germany
SuperMUC-NG is at No. 9 position with 19.5 petaflops. The Lenovo-built machine is powered by Intel Platinum Xeon processors.
Lassen supercomputer, installed at Lawrence Livermore National Laboratory in the United States
Lassen supercomputer is ranked at No. 10 and delivers 18.2 petaflops. It is said to be the unclassified counterpart to the classified Sierra system and shares the same IBM Power9/NVIDIA V100 GPU architecture.
Indian supercomputers in the list: Pratyush, installed at Indian Institute of Tropical Meteorology in Pune
Pratyush is a Cray XC40-based super computer. It is ranked at No. 57 in the Top 500 list.
Indian supercomputers in the list: Mihir, National Centre for Medium Range Weather Forecasting, Noida
Ranked at No. 100 in the overall list, Mihir supercomputer is another Cray XC40-based system.
As in 2020
Every year the list of fastest supercomputers comes out. Published by TOP500, the list this year features two supercomputers from India once again. Pratyush and Mihir, the two supercomputers of India, ranked in at 67th and 120th spot on the list. These rankings were announced on June 22 at the ongoing virtual event ISC (International Supercomputing Conference) High Performance 2020 Digital.
The two Indian supercomputers’ rankings, however, has fallen down compared to last year. According to the 2019 rankings, Pratyush supercomputer was in 57th spot whereas Mihir supercomputer’s ranking was 100. The TOP500 list is a project that regularly ranks and evaluates the top 500 fastest supercomputer systems in the world.
Prathush is India’s fastest supercomputer with 3.7 petaflops high performance. The supercomputer is at the Indian Institute of Meteorology, Pune.
For the uninitiated, a petaflop is the measure of a computer’s processing speed. A petaflop is equal to a quadrillion or a thousand trillion calculations per second.
Mihir, which was ranked 120th, had a capacity of 2.5 petaflops. This particularly fast machine is at the National Centre for Medium Range Weather Forecast in Noida.
The two supercomputers were installed in 2018 by the ministry of earth sciences. The estimated cost of installing and setting them up was close to Rs 440 crores, according to an earlier report by The Economic Times.
The two supercomputers are used to help the quality of not just weather forecasts like monsoon but also of cyclones, earthquakes and other extreme events. They can also help forecast air quality, flood, drought among other things.
As for the fastest supercomputer in the world, that title was accorded to Fukagu, a Japanese system backed by Fujitsu and RIKEN. For comparison sake, Fukagu had a performance of 415.53 petaflops. The list, otherwise, was dominated by a lot of supercomputers from China and the USA.
2022: Param Pravega
Under the National Supercomputing Mission (NSM), the Indian Institute of Science (IISc) Bengaluru has installed and commissioned Param Pravega, one of the most powerful supercomputers in the country. The NSM was launched to enhance the research capacities and capabilities in the country by connecting them to form a Supercomputing grid, with National Knowledge Network (NKN) as the backbone. The NSM is setting up a grid of supercomputing facilities in academic and research institutions across the country. Here are key details about Param Pravega supercomputer and more.
What is the hardware of Param Pravega supercomputer
The Param Pravega supercomputer is a mix of heterogeneous nodes built of Intel Xeon Cascade Lake processors for the CPU nodes and NVIDIA Tesla V100 cards on the GPU nodes. It consists of 2 Master nodes, 11 Login nodes, 2 Firewall nodes, 4 Management, 1 NIS Slave and 624 (CPU+GPU) compute nodes with total peak computing capacity of (CPU+GPU) 3.3 PFLOPS performance. The hardware of Param Pravega is an ATOS Bull Sequana XH2000 series system, boasting a comprehensive peak compute power of 3.3 PetaFlops.
What is the operating system of the Param Pravega
The entire system is built to operate using Linux OS based on CentOS 7.x distribution. The machine hosts an array of program development tools, utilities, and libraries for ease of development and execution of HPC applications on the heterogeneous hardware of the machine. Common compilers from GNU and Intel are accessible on the system for MPI and OpenMP parallel libraries. For use on the GPU nodes the system has CUDA and OpenACC SDKs installed. Further, popularly used parallel mathematical, scientific, and application libraries like Intel-MKL, GNU Scientific library, HDF5, NetCDF, range of Python-based mathematical and data manipulation libraries etc. are installed on the system. Added to these, the machine also hosts system monitoring and management tools developed by CDAC team.
Who provides design and implementation support
The software stack on top of the hardware is provided and supported by C-DAC, Pune. The HPC Technologies team of C-DAC is instrumental in the design and implementation of the solution for end-use along with ATOS and BULL.
Where are the other supercomputers installed in India
NSM has supported the deployment of 10 supercomputing systems so far in the country. These include several IITs; Indian Institutes of Science Education and Research (IISER), Pune; Jawaharlal Nehru Centre for Advanced Scientific Research (JNCASR), Bengaluru; National Agri-Food Biotechnology Institute (NABI), Mohali; and C-DAC Pune. PARAM Shivay, the first supercomputer assembled indigenously, was installed in IIT (BHU), followed by PARAM Shakti, PARAM Brahma, PARAM Yukti, PARAM Sanganak at IIT-Kharagpur IISER, Pune, JNCASR, Bengaluru and IIT Kanpur respectively. Four NSM Nodal Centres for training in HPC and AI have been established at IIT Kharagpur, IIT Madras, IIT Goa and IIT Palakkad.
PARAM Siddhi supercomputer made it to 'Top 100 supercomputers' in the world list in 2020
PARAM Siddhi - AI, the high-performance computing-artificial intelligence (HPC-AI) supercomputer, has achieved global ranking of 62 in TOP 500 most powerful supercomputer systems in the world, released in November 2020.
Is this the first supercomputer at IISc Bengaluru
No. IISc has a cutting-edge supercomputing facility which was established several years ago. In 2015, the institute procured and installed SahasraT, which was at that time the fastest supercomputer in the country.
What all is SahasraT supercomputer used for
The SahasraT supercomputer hosts an array of programme development tools, utilities, and libraries for developing and executing High Performance Computing (HPC) applications. Faculty members and students have been using this facility to carry out research in various areas. These include research on Covid-19 and other infectious diseases, such as modelling viral entry and binding, studying interactions of proteins in bacterial and viral diseases, and designing new molecules with antibacterial and antiviral properties. Researchers have also used the facility to simulate turbulent flows for green energy technologies, study climate change and associated impacts, analyse aircraft engines and hypersonic flight vehicles, and many other research activities. These efforts are expected to ramp up significantly with Param Pravega.