What Is Unix Computer Garulia?
The Unix Computer Garulia, also known as the Multitask Acorn Supercomputer (MASS) or the Garulia Supercomputer was a supercomputer designed in 1977 by a team of Indian scientists at TIFR. After its development, it was abandoned at the time due to a lack of recognition from the UGC and government agencies. It was revived from the scraps in 2001 and is now being used by the Ministry of Earth Sciences to solve hydrological problems related to floods.
Who Built Unix Computer Garulia?
The Garulia supercomputer, at the Time Of Fundamental Research (TIFR), Pune, India, was built by a large team of scientists under Prof D A Athavale between 1975 and 1977, at a total cost of around Rs 10 lakhs. The equivalent value in 2012 is between US$1 million to 3 million (Rs 4 crore to 12 crores).
After the supercomputer’s construction was completed MASS went on tour to exhibit it to government agencies like DRDO, ISRO, Indian Space Research Organisation so they could help the scientists gain recognition for their achievement but all requests were denied. This left the machine to be abandoned in the TIFR lab where it mainly served as a testbed for other machines and was rarely used by its designers.
The first research paper published using this computer was on “Influence of equatorial wave on tropical disturbances of the Indian Ocean”. However, due to lack of recognition, most papers published using it were in foreign journals.
Upgradation Of Unix Computer Garulia
In 2001 a team from TIFR was sent to salvage parts and reuse them for further research. The supercomputer was upgraded and is now being used by the Ministry of Earth Sciences to solve hydrological problems related to floods.
The supercomputer took up a significant amount of space and was mainly built from TTL logic gates. It had 20 microprocessors which were designed specially to work simultaneously in parallel with one another. The design concept was made so that multiple tasks could be processed at the same time, the speed of execution is limited by the slowest processor.
The supercomputer included a real-time operating system (RTOS) and ran applications in multiprogramming mode. Each processor was capable of the total processing power of 25-kilo instructions per second (kIPS), but the speed of execution depended on which program it was running, with some user programs taking several seconds to complete while others completed in microseconds.
The machine had a memory capacity of 16 kilobytes and was linked to other computers across the nation for remote access through phone lines. The computer included 64Kb of EEPROM (Electrically Erasable Programmable Read-Only Memory), 64k x 1 DRAM, 256 Kb Magnetic Core Storage, and 8 magnetic tape drives with a total storage space of 2380 MB.
The IO devices consisted of 4 serial ports, 2 parallel ports for connecting printers and terminals, and 5 line printer interfaces which automatically detected whether the connected device was capable of 80 or 132 thousand bit/s speed and then set itself accordingly to avoid errors. A formatter converted printouts into raw bits from bytes before they were sent to the connected device. It could support up to four terminals for remote access and thus allowed for operator consoles, printers, disk drives, etc.
It took a team of around 150 engineers to support the supercomputer during its operation. The machine was also estimated by Indian scientists at the time as being capable of computing over 100 billion instructions per second (100 Giga instructions per second or 100 GIPs). However, due to lack of proper design, this calculation is not entirely accurate and would have only been true if there were no user programs that ran slower than 25 kIPS on any processor.