Chapter: Requirment of the Super computer
Date: March 12, 1997
The Prime Minister's Office conference room was packed with senior government officials, senior scientists, and industry titans of India's IT sector. Prime Minister Animesh, his face a mask, sat at the far end of the long table, his fingers following the thickness of the file in front of him.
In front of him were the Minister of Science & Technology, the Finance Minister, and the cream of C-DAC (Centre for Development of Advanced Computing), ISRO, and DRDO—the premier scientific research organizations of India. The silence was heavy in the room, broken occasionally by the distant whir of the air conditioner.
The Science Minister tapped his foot restlessly before he spoke at last.
"Sir, why have we been summoned today?"
.
Animesh leaned forward, his sharp eyes on the Science Minister. He opened the folder and pushed it down the table.
"This," he said softly but firmly, "is India's future."
The officials opened the file, their eyes growing wide as they read what was inside. It was a simple blueprint for India to purchase or develop at least 50 supercomputers over the next year or so before the year 2000.
"Supercomputers, Prime Minister?" one of the C-DAC scientists asked. "We are developing PARAM, but this magnitude."
Animesh nodded. "Yes. And we're not proceeding as quickly as we must. India can't lag behind in computers. I want India to have at least 50 supercomputers by the year 2000, made here or purchased."
The corridor was filled with hushed tones. Most of the officials looked around at each other anxiously, and some of them started flipping through the pages of the file, reading the opulent schedule.
"Excuse me, sir," said the Minister of Science, "supercomputers are an enormous, enormous expense. The new US Crays are running several hundred million dollars. Even constructing PARAM took years. Doubling it to 50 will be."
Animesh raised a hand, silencing the demur.
"I see the issues. That is why I have called for this meeting—to discover solutions, not excuses."
Why India Needs Supercomputers
He stood up from his chair and approached the large world map on the wall.
"See the world stage," he said, stretching out his hand in the direction of the United States, Japan, and Europe. "They have the finest in supercomputer power, propelling their defense systems, their weather forecasting, their industrial research, and economic modeling. And what do we have?".
How are we even going to match up in space exploration with NASA when they are using computers to do millions of simulations a day and we are doing small calculations in weeks?
He addressed DRDO.
"How are you going to create next-generation weapons, missile systems, and defense simulations when you don't have the brute computing power to do heavy-duty calculations?"
Then he approached the Finance Minister.
"And how are we going to plan for our economic growth when we cannot even handle real-time production, market, and trading data?"
Silence. Caught everyone off guard with how grave it all was.
A Two-Phase Plan
Animesh reclined and clasped his fingers together.
"We shall work in two phases."
Phase 1: Immedidate Acquisition (1997-1998)
India will negotiate the purchase of at least 10-15 high-end supercomputers from the US, Japan, or Europe on urgent needs.
ISRO, DRDO, Finance Ministry, C-DAC, and Meteorological Department are the priority segments.
India will utilize diplomatic efforts and trade negotiations to acquire the systems on best possible terms.
Phase 2: Indigenous Development (1998-2000)
C-DAC and Indian IT sectors will be completely funded to accelerate indigenous supercomputer development.
Public-private partnerships would be established to involve superior Indian IT companies for research and development.
India must have at least 50 functioning supercomputers by the year 2000, of which at least 30 must be indigenous.
The Finance Minister cleared his throat. "Sir, this will cost. We've already allocated money for economic liberalization and infrastructure building. Where do we find room for this in our budgets?"
Animesh smiled faintly. "Tell me, Minister, how much does it cost to lag behind?"
He paused dramatically before continuing.
"If we don't invest now, we will be dependent on foreign nations for computer capability, just as we were dependent on foreign nations for fighter jets, satellites, and even telephone installations. The price of missing out on supercomputers will be many times the price of the investment now."
He addressed the Science Minister. "And what if the US won't sell?"
The minister hesitated. "There are other nations, sir. Japan, Germany, even Russia."
Animesh nodded. "Try all channels. But we cannot lose. If none sell, we produce. But we will have 50 supercomputers in 2000."
There was quiet this time in the room, but no doubt—only determination.
Animesh rested back in his chair and drummed on the table. "I want proposals from all departments within seven working days' time. The budgets will be sanctioned forthwith. India is globalizing, and the world is digital. India's not going to be left behind."
The meeting had turned somber. Officialdom and ministerial faces exchanged glances, sensing the gravity of what the Prime Minister was stating. India required supercomputers, and India required them in all urgency.
As the murmurs were soft, Animesh knocked his knuckles on the table again. "Let us think this on a sane level. I need information. No ifs and buts. Action is needed."
Reaction of C-DAC: Developing the Next PARAM
C-DAC's Director and one of the very few individuals in the room who were not alive or nervous responded.
Sir, we already developed PARAM 8000 and its subsequent version PARAM 9000. We wanted to reach the level of 100 GFLOPS (gigaflops), but once more we are lagging behind the US and Japan, who have already moved on to the level of teraflops.
Animesh nodded. "I know. When can you upscale?
The Director took a deep breath. "If we get the total funding, we can step up PARAM's next-generation model—maybe to a teraflop—by 1999. But we will require unrestricted access to hardware and software, including processors."
The Prime Minister addressed the Finance Minister. "What's the estimate?"
"To create an indigenous system to equal the world, we would need at least $250 million to $300 million in the next three years."
A nod, and Animesh nodded in affirmation. "Approved."
The Finance Minister blinked. "Sir, like that?"
Animesh leaned forward. "India waits for no man. C-DAC will be funded to the hilt. This is no luxury—it's a necessity."
The ISRO Chairman, Dr. Krishnaswamy, pushed his spectacles up before speaking.
"Sir, we need supercomputers to compute the orbits of satellites, forecast the weather, and simulate deep space. We currently have some small computers, but we can't compete abroad without an actual supercomputer. America and Europe already possess special systems for their space agencies."
Animesh nodded. "What do you need, Doctor?"
"We need five working machines simultaneously, if possible from a healthy foreign source. We'll have to wait if we manufacture ourselves."
The Prime Minister turned to the Foreign Minister. "Go and talk to Japan and France. If the US will not sell, we look elsewhere. Tell them categorically that ISRO is willing to cooperate in space research in lieu."
Foreign Minister replied this. Director. "Alright, sir. I will make a contact with NEC (Japan) and Bull (France)."
DRDO Requirements: Defense Simulations & Weapons Research
Dr. APJ Abdul Kalam, the head scientist of DRDO (Defense Research and Development Organization), spoke next. He had been listening attentively, and now he put a folder on the table.
"Prime Minister, in the absence of supercomputers, our nuclear tests, missile research, and defense programs for the future will be gravely impaired. We are working with outdated hardware today, and we take weeks to do something America and China do in hours."
Animesh leaned forward. "What do you need most urgently?"
Dr. Kalam did not waste time. "We require a minimum of ten high-performance computing clusters with expertise in missile technology, aerodynamics, and nuclear simulations. If we are able to lease foreign supercomputers on an emergency basis, we can proceed with the indigenous ones in parallel."
The Prime Minister addressed the Finance Minister. "Fund it. We will invest another $500 million into DRDO's computer infrastructure within the next three years."
A Controversial Proposal: To Buy or Not to Build
The Science & Technology Minister raised a point of concern. "Sir, I am going with my heart as I am okay with this move, but aren't we relying too much on foreign supercomputers? If we are planning to purchase them, there is always a possibility that they have backdoors in them or restriction on their use for the army.".
Animesh smiled. "That is why we are doing both- buying to meet short-term needs and building for long-term autonomy. C-DAC will expand on indigenous platforms, while ISRO and DRDO will acquire the equipment they need in the beginning."
The Minister nodded, reassured by equilibrium.
Telecom Minister Speaks: Supercomputers & The Internet
By this time, the Telecom Minister had interjected.
"Sir, if we are spending on supercomputers, then we need to spend on the development of our communication infrastructure as well. Our backbone on the internet is poor as it is, and barely a handful of dozen institutions have high-speed connectivity to them."
Animesh asked him to continue.
"By 2001, we must have high-speed internet connectivity in all our universities, research institutions, and government ministries. Otherwise, however much we construct good solid supercomputers, they will starve on bad connectivity."
The Prime Minister's eyes sparkled. "That is a wonderful proposal. What would it cost us likely?"
"To establish a national high-speed network, we would need to spend a minimum of $1 billion. This would include fiber-optic upgradation, satellite links, and dial-up to high-speed internet upgrading."
Animesh glanced at the Finance Minister. "Allocate a separate budget for telecommunication infrastructure. This will run parallel to our supercomputer project."
Once all departments had presented, Animesh laid out the most critical decisions:
India will acquire a minimum of 50 supercomputers by the year 2000—foreign acquisition as well as indigenous development.
C-DAC will be given $300 million to design its next-generation PARAM series.
ISRO will be equipped with five high-performance foreign-supplied supercomputers, and formal negotiations will be initiated forthwith.
DRDO will be given $500 million for computational infrastructure.
A country-wide roll-out of broadband internet will be funded with an investment of $1 billion.
Foreign hardware would be brought in through diplomatic avenues as the country builds domestic ones.
Animesh brought his hands together and looked out over the room.
The meeting was concluded. The officials left, energized by the meeting.
The sun was setting outside of Delhi, but within the Office of the Prime Minister, the future had arrived for the technological life of India.