Indian advancements in Supercomputing

[nrj]

well there is no problem if india goes for nuclear test now apart from a little fuss from china,pakistan,america,eu,nsg but the problem is that after signing the nuclear deal with US manmohan singh has entangled our strategic nuclear programme in a spiders web.many country will impose restictions on us and the civilian nuclear sector depends heavily on uranium imports until india perfects the thoriunm cycle.

Apparently right.

But very little reason to go for physical test after 20 years when our supercomputers can validate the same.

 

[SHASH2K2]

Apparently right.

But very little reason to go for physical test after 20 years when our supercomputers can validate the same.

Nrj question here is that do we have sufficient data to simulate a thermo nuclear bomb ? If experts are correct then thermo nuclear bomb didn't explode properly and in that case data used for simulation may not be correct .

 

[nrj]

Nrj question here is that do we have sufficient data to simulate a thermo nuclear bomb ? If experts are correct then thermo nuclear bomb didn't explode properly and in that case data used for simulation may not be correct .

Right. And it is where we arrive on 2 possibilities -

1. Explosion wasn't proper & collected data is insufficient.

In this case, no doubt attempted simulation will give inferior results. This puts added task to conduct another physical test.

2. Explosion was proper & collected data is accurate.

We are/were green to go for simulation exercise & build relative judgments.

--

It was inferred that GOI had accurate test last time but over the period of time, Security advisers felt the need to re-test & put the final end on required offensive experiments so it was made out in public that Last test wasn't proper & couldn't gather sufficient data.

Many of think-tanks still feel that this was a gimmick in order to find excuse paving way for another Indian Nuclear test.

Whichever maybe the case, I think it's going to take some time for actual information to appear in public domain.

 

[SHASH2K2]

Right. And it is where we arrive on 2 possibilities -

1. Explosion wasn't proper & collected data is insufficient.

In this case, no doubt attempted simulation will give inferior results. This puts added task to conduct another physical test.

2. Explosion was proper & collected data is accurate.

We are/were green to go for simulation exercise & build relative judgments.

--

It was inferred that GOI had accurate test last time but over the period of time, Security advisers felt the need to re-test & put the final end on required offensive experiments so it was made out in public that Last test wasn't proper & couldn't gather sufficient data.

Many of think-tanks still feel that this was a gimmick in order to find excuse paving way for another Indian Nuclear test.

Whichever maybe the case, I think it's going to take some time for actual information to appear in public domain.

If you check this link first possiblity is much much stronger. K santhanam is not a man that can be easily rubbished.

Pokhran-II Thermo-Nuclear Test: A Failure

 

[EagleOne]

Based on this evidence the most plausible interpretation of the 11 May tests is that the fission bomb was successful and yielded 12 kt. The straightforward nature of this technology and the lengthy time India had to master it makes doubts about its performance difficult to support. The yield of the thermonuclear device was apparently 22-25 kt. This is at or above the expected yield range of a thermonuclear primary, so it appears that the fusion boosted device used for the primary was successful. The plausible yield for the secondary stage is from 10 to 15 kt, well below the 30 kt or so apparently expected by India. The 0.3 kt yield for the third device is appropriate for the unboosted yield of a boosted fission design - that is, a device that is fired without the deuterium-tritium boost gas. It suggests that India now as the ability to employ full yield fusion boosted weapons using less-than-weapon grade plutonium. This could be fuel grade plutonium as is produced by India's power reactors or perhaps an intermediate grade. In the first case all of India's considerable holdings of plutonium could be used for weapon manufacture, in the second case the stock of weapon grade material could be extended by mixing it with fuel grade plutonium.

The reasons given for the second group of tests are entirely credible and are consistent with the objectives of other nuclear weapons nations in recent testing activities (either nuclear tests prior to the CTBT signing, or sub-critical tests afterward). India's stated interest in gathering high quality physical data for use in simulations is quite consistent with a well planned weapon development effort.

The use of simultaneous detonations of multiple nuclear devices has been a routine practice by other nuclear weapons states. The principal purposes of this technique are to minimize the ability of other nations to collect intelligence data about the tests (ironically too successful, given India's subsequent and not entirely convincing efforts to publicize its test results), and to reduce the cost and difficulty of test preparations. In India's case the desire to minimize political repercussions and outside pressure also motivated a compressed schedule of simultaneous tests.

Corroboration of a sort for India's explanations for the different tests on 11 May was given immediately after they were fired, and four days before the official AEC/DRDO announcements, by commentary on their military significance offered by P.K. Iyengar, the former chairman of India's AEC. On 12 May he told Reuters in Mumbai (Bombay) that their differing sizes corresponded to three ways in which nuclear bombs might be used.

According to Iyengar, the smallest was the size that might be fired as an artillery shell or dropped from a combat support aircraft. The mid-size blast was from a standard fission device equivalent to about 12 kilotons -- the size that might be dropped from a bomber plane. The largest of the three warheads tests on Monday was not a full hydrogen bomb. Most of its 50 to 100 kiloton explosive force came from the primary, a fission device which serves as a trigger for the H-bomb's big fusion explosion. Iyengar said the device contained only a token amount of the hydrogen variant tritium. It showed that India's thermonuclear technology worked, but did not produce the megaton explosion typical of a full H-bomb.

"We need not go for a megaton explosion while testing an H-bomb," said Iyengar, one of the scientists involved in India's only other nuclear test, in 1974. "Such tests are required only if we are planning for a total destruction of the opposite side. They don't have relevance in our strategy."

One technical issue that affects all three of the low yield tests, with yields of 0.2, 0.3, and 0.5 kilotons, is that the yields of low yield fission tests are very sensitive to physical parameters like the amount of fissile material present, the degree and uniformity of compression, and the nuclear properties of the materials. Even if the Indian designers have achieved very accurate control over the implosion process, without test data in this yield range to calibrate their models they would have difficulty predicting the exact yield of their devices. This of course is one reason for conducting these tests in the first place, but it also means that there is a good chance of overshooting or undershooting the target yields. Thus the intended yields of these devices may be significantly different than the yields produced, on the other hand if the advertised yields are the expected ones, then the actual ones may be significantly different.

and also drdo and barc scientist asked for approval for doing a TN test in 2003 which was not approved
they may want to do this because of disputed yeild .is still a question .?

and ironically mr santanam or some one will only talk about the yeild when proliferation discussions are going or some thing like that
-caution

so according to present situation cong govt is not going for any nuclear test in near future up 2014
so after that is there any posibilty? with present nsg waiver and 123 agrement in place.
actually 123 will void if we conduct a nuclear test

 

[LETHALFORCE]

Nrj question here is that do we have sufficient data to simulate a thermo nuclear bomb ? If experts are correct then thermo nuclear bomb didn't explode properly and in that case data used for simulation may not be correct .

confidence in simulation capability

 

[LETHALFORCE]

http://economictimes.indiatimes.com...fastest-supercomputer/articleshow/8144400.cms

ISRO builds India's fastest supercomputer

BANGALORE: Indian Space Research Organisation has built a supercomputer, which is to be India's fastest in terms of theoretical peak performance of 220 TeraFLOPS (220 Trillion Floating Point Operations per second).

The supercomputer "SAGA-220", built by the Satish Dhawan Supercomputing Facility located at Vikram Sarabhai Space Centre (VSSC), Thiruvananthapuram at a cost of about Rs 14 crore was inaugurated by K Radhakrishnan, Chairman ISRO at VSSC today, ISRO said in a statement.

The new Graphic Processing Unit (GPU) based supercomputer, "SAGA-220" (Supercomputer for Aerospace with GPU Architecture-220 TeraFLOPS) is being used by space scientists for solving complex aerospace problems.

"SAGA-220" is fully designed and built by Vikram Sarabhai Space Centre using commercially available hardware, open source software components and in house developments.

The system uses 400 NVIDIA Tesla 2070 GPUs and 400 Intel Quad Core Xeon CPUs supplied by WIPRO with a high speed interconnect.

With each GPU and CPU providing a performance of 500 GigaFLOPS and 50 GigaFLOPS respectively, the theoretical peak performance of the system amounts to 220 TeraFLOPS, the statement said.

The present GPU system offers significant advantage over the conventional CPU based system in terms of cost, power and space requirements, it said.

The system is environmentally green and consumes a power of only 150 KW. This system can also be easily scaled to many PetaFLOPS (1000 TeraFLOPS).

 

[chex3009]

^^^Where it stands in worldwide standings???

 

[LETHALFORCE]

^^^Where it stands in worldwide standings???

http://business.rediff.com/slide-sh...-tech-top-10-supercomputers-none-in-india.htm

 

[pmaitra]

If you check this link first possiblity is much much stronger. K santhanam is not a man that can be easily rubbished.

Pokhran-II Thermo-Nuclear Test: A Failure

Some excerpts from the article:

(b) design and validate new weapons using our powerful indigenous computers—without having to undertake further field tests.

Indigenous? How? India's most recent and powerful computer SAGA-220 has vital foreign components. Do we have any commitments with the manufacturers that we will not be able to use them for nuclear testing?

Secondly, our "clean" NSG waiver enables us to undertake unrestricted nuclear trade with all countries. Our testing will only lead to termination of the US 123 Agreement and some easily faceable civil sanctions by Congress.

NSG waiver was given with the understanding that India will not conduct tests. Waiver can always be revoked.

Said that:

• We need a lot more testing data to have confidence in our design. Two conditions must be satisfied: Firstly, we have to have a large volume of data , and secondly, the data must pass the T-test and F-test and several other methods of testing to be sufficient and statistically significant.
• We need to evaluate the worst-case scenario if we test more T-nukes. If they have to be carried out, they better be carried out now than later.

 

[LETHALFORCE]

Some excerpts from the article:



Indigenous? How? India's most recent and powerful computer SAGA-220 has vital foreign components. Do we have any commitments with the manufacturers that we will not be able to use them for nuclear testing?

Interesting point Pmaitra. How would the manufacturers know if we are using their chips for nuclear test simulations??

 

[pmaitra]

Interesting point Pmaitra. How would the manufacturers know if we are using their chips for nuclear test simulations??

True. They needn't know. The caveat is we do not make any computer that does not have any vital components that are not imported. Which means, we simply cannot perform supercomputing indigenously. Nonetheless, no one needs to know how we did our computing anyway.

 

[prahladh]

Some excerpts from the article:



Indigenous? How? India's most recent and powerful computer SAGA-220 has vital foreign components. Do we have any commitments with the manufacturers that we will not be able to use them for nuclear testing?

The same way Tejas or other projects are.

 

[pmaitra]

The same way Tejas or other projects are.

That is exactly my point. One example would be the complications India is facing trying to export HAL ALH to Myanmar.

 

[A.V.]

India's fastest supercomputer Built by ISRO

BANGALORE: Indian Space Research Organisation has built a supercomputer, which is to be India's fastest in terms of theoretical peak performance of 220 TeraFLOPS (220 Trillion Floating Point Operations per second).

The supercomputer "SAGA-220", built by the Satish Dhawan Supercomputing Facility located at Vikram Sarabhai Space Centre (VSSC), Thiruvananthapuram at a cost of about Rs 14 crore was inaugurated by K Radhakrishnan, Chairman ISRO at VSSC today, ISRO said in a statement.

The new Graphic Processing Unit (GPU) based supercomputer, "SAGA-220" (Supercomputer for Aerospace with GPU Architecture-220 TeraFLOPS) is being used by space scientists for solving complex aerospace problems.

Economic times report

 

[The Messiah]

This is a good development.

How is it comparable to other super computers in terms of performance ?

 

[badguy2000]

This is a good development.

How is it comparable to other super computers in terms of performance ?

its speed is about 1/10 of the global fastest super computer.

 

[rudresh]

its speed is about 1/10 of the global fastest super computer.

nonsense

it is one of the milestones where india used graphics processing to improve speed.........fyi
today any country can build a super computer its advantage lies with its power consumption to its number crunching ability.

 

[JAISWAL]

Rs 10,000-crore push for India's supercomp plan

Rs 10,000-crore push for India's supercomp plan

India this week committed Rs 10,000 crore to indigenously develop the world's fastest supercomputer by 2017. The Planning Commission agreed in principle to provide the funds to the Indian Space Research Organsiation (ISRO) and Indian Institute of Science (IIS), Bangalore to develop a supercomputer with a performance of 132.8 exaflops (132 quintillion floating operations per second). A quintillion has 18 zeros (a million has six).

In computing, flops is an acronym to measure computing performance. An average personal computer can go up to 7.5 gigaflops

The world's fastest supercomputer right now is a Chinese one, which can do 2.7 pentaflops, or two quadrillion flops. A quadrillion has 15 zeros
.
India in 2007 had the world's fourth fastest indigenously-developed supercomputer with a performance of 172.5 teraflops (172 trillion flops), which has been enhanced this month to 220 teraflops. That's still a level lower than China's supercomputer.

The Indian supercomputer will not be used only for enhancing the country's space abilities, it will also be used to predict monsoon and precise weather inputs to boost agriculture N Balakrishnan, associate professor at IIS-Bangalore, said the target being set is "ambitious" while referring to achieving the exaflop – or next level of computing performance -- by 2017. "We have planned everything minutely."

"We have agreed to provide R10,000 crore for having 'exa' level of supercomputing facility," minister of state for planning Ashwini Kumar told HT.

Balakrishnan, in a presentation to the plan panel, said ISRO has already booked key equipment to develop the supercomputer by 2017. "Most of the other gadgets will be indigenously developed," he said. "Supercomputing is key to competing in the international space market," Balakrishnan said.

LINK--
Rs 10,000-crore push for India

 

[p2prada]

132800 petaflops as compared to China's 2.7 petaflops. 6 years is indeed ambitious.