4 supercomputers from India in world's fastest 100

[Guynextdoor2]

Defence News - India To Make World's Fastest Super Computer

I love this arrangement. IISC has an M Tech in super computing sciences. So the kids will get to work on this even when they're studying. Best way forward- build the machine and build the talent pool.

I dont know why cheerleaders are still saying lack of funds when they are trying to pump $2 billion into this.
http://www.defence.pk/forums/central-south-asia/259243-india-will-have-super-supercomputer-2017-a.html

Where is the money coming from all of a sudden?

india has a chronic shortage of electric power, flooding etc. How do you cope with those?

No one asked you for advice on policy so shut it

 

[shuttler]

No one asked you for advice on policy so shut it

India too has committed $ 2 billion. Indian Space Research Organization (ISRO) and the Indian Institute of Science Bangalore are expected to develop a supercomputer with 132.8 exaflops by 2017.

Source: http://www.defence.pk/forums/centra...super-supercomputer-2017-a.html#ixzz2WlhQPbDz

You have earmarked $2 billion for a "132.8 exaflops by 2017". Has the budget allowed anything for backup diesel generators, cooling system, drainage system? Is "exaflops" or "2017" or both typos for "petaflops and 2071"?

 

[Guynextdoor2]

You have earmarked $2 billion for a "132.8 exaflops by 2017". Has the budget allowed anything for backup diesel generators, cooling system, drainage system? Is "exaflops" or "2017" or both typos for "petaflops and 2071"?

Yeah, why don't you focus on your own drainage for now. Last time I checked they were turning your rivers red because of unchecked pollution and 'hundreds' of chinese were floating up our rivers despite mother china being sooo low on crime. Maybe those poor guys need your attention there than on our poor drainage systems.

 

[buddyboyyash]

Lack of funds？

You do know how to find excuses for yourself。

A low petaflop computer costs less than 100 million dollars now a days。

You are buying helicopters worth tens of millions of dollars each from the US，yet you don't have the money to build ONE petaflop computer？

now....let me explain you something.....

FUNDS in this context means the amount of money being invested in supercomputing research.....

the funds allocated for research were insufficient to make a world beating supercomputer (which we are talking about here)....

and as far as i understand...we are talking about the capability of supercomputers...and not the number of supercomputers installed in the country

so....your statement of "a low petaflop comp costing 100 million" might be true but very insignificant here

 

[shuttler]

now....let me explain you something.....

FUNDS in this context means the amount of money being invested in supercomputing research.....

the funds allocated for research were insufficient to make a world beating supercomputer (which we are talking about here)....

and as far as i understand...we are talking about the capability of supercomputers...and not the number of supercomputers installed in the country

so....your statement of "a low petaflop comp costing 100 million" might be true but very insignificant here

and you think you guys can come up with anything "world beating" like your claim with $2 billion now and why not before?
if that is the case, that only means your government fails to have a long term vision and you are implementing policies based on knee jerk reflexes

 

[Skull and Bones]

I hear India has a big IT industry. Are there any indigenous processors being developed within the country that could be utilised for supercomputers any time soon?

Having a developed IT industry doesn't necessarily mean one needs to have an advanced semiconductor processing industry and fabrication facility. And yes, India sucks in the latter part.

 

[buddyboyyash]

and you think you guys can come up with anything "world beating" like your claim with $2 billion now and why not before?
if that is the case, that only means your government fails to have a long term vision and you are implementing policies based on knee jerk reflexes

well unfortunately...yes...this govt does fail to have a long term vision....completely agreed on this point

didnt really understand your first sentence

 

[anant_s]

I love this arrangement. IISC has an M Tech in super computing sciences. So the kids will get to work on this even when they're studying. Best way forward- build the machine and build the talent pool.

yes it is vitally important to have guys trained for these high end systems or learn to use these as tools for furthering their science. It is an open fact that we donot have a good enough semi-conductor industry but then how many nations have them and you cannot sit around waiting for it to develop. There are enough resources available in market, make smart use of them.
btw the hardware we are using to assemble our supercomputers are available to everyone in the world but the fact that very few countries are able to commission and make use of such systems does tell that we are doing things right.

 

[Guynextdoor2]

yes it is vitally important to have guys trained for these high end systems or learn to use these as tools for furthering their science. It is an open fact that we donot have a good enough semi-conductor industry but then how many nations have them and you cannot sit around waiting for it to develop. There are enough resources available in market, make smart use of them.
btw the hardware we are using to assemble our supercomputers are available to everyone in the world but the fact that very few countries are able to commission and make use of such systems does tell that we are doing things right.

semi conductor industry has not developed primarily because of power related issues, not talent related ones. Even if there is a loss of 15 mins of power, the loss could be in millions.

Having a developed IT industry doesn't necessarily mean one needs to have an advanced semiconductor processing industry and fabrication facility. And yes, India sucks in the latter part.

power infra issue- else many ppl are keen on setting that here

 

[bloo]

and you think you guys can come up with anything "world beating" like your claim with $2 billion now and why not before?
if that is the case, that only means your government fails to have a long term vision and you are implementing policies based on knee jerk reflexes

We are not rich exhibitionists like the Chinese and $2 Billion is a big sum. Just coz you are probably hearing the news now doesn't mean that the plans weren't there for a long time, all these decisions go thru a budget meeting and are not knee jerk reflexes.

Why did China suddenly want to make J-20 or any such 5th gen plane? Was it based on a knee jerk policy or was it just to showoff?

Why don't you just come out and say it that an article praising India by even a minute figure just pisses you off?

 

[Bobby]

We are 2 trillion economy 11 supercomputer is enough for now....we are not in any race ....

we are tortoise.... let the rabbit run fast.....

 

[Areesh]

None in the top 500.

I don't get it. Why Pakistan isn't in top 500. Shouldn't NUST's super computer be in top 500 with a peak speed of 132 tera flops?


Nust launches Pakistan’s fastest supercomputer – The Express Tribune

 

[shuttler]

well unfortunately...yes...this govt does fail to have a long term vision....completely agreed on this point didnt really understand your first sentence

The first sentence of my post read like:

and you think you guys can come up with anything "world beating" like your claim with $2 billion now and why not before?

and your government's recent relevation of this investment:

India too has committed $ 2 billion. Indian Space Research Organization (ISRO) and the Indian Institute of Science Bangalore are expected to develop a supercomputer with 132.8 exaflops by 2017.

Source: http://www.defence.pk/forums/indian...dia-worlds-fastest-100-a-4.html#ixzz2WmJJOGUv

June issue of the Top500 showed india's best supercomputer is at the teraflop level. Is it realistic to have 132.8 exaflops by 2017? If you claim you can do it in 4 years' time beating the rest of world, why not start doing it 4 years ago or earlier! - that was what I meant

 

[shuttler]

We are not rich exhibitionists like the Chinese and $2 Billion is a big sum. Just coz you are probably hearing the news now doesn't mean that the plans weren't there for a long time, all these decisions go thru a budget meeting and are not knee jerk reflexes.

it is you who declared the $2 billion spending on a 132 exaflops. Where the rest of the world has declared that? Now who is the exhibitionist?

Why did China suddenly want to make J-20 or any such 5th gen plane? Was it based on a knee jerk policy or was it just to showoff?

We have planned our fifth gen fighter jets since 2002, first flight took off 2011
You claim your exaflop in 4 years time while at present you top dog does not exceed 1 petaflop

Why don't you just come out and say it that an article praising India by even a minute figure just pisses you off?

You will get your praise when you have earned it!

 

[buddyboyyash]

The first sentence of my post read like:


and your government's recent relevation of this investment:



June issue of the Top500 showed india's best supercomputer is at the tetraflop level. Is it realistic to have 132.8 exaflops by 2017? If you claim you can do it in 4 years' time beating the rest of world, why not start doing it 4 years ago or earlier! - that was what I meant

its India man....thats my answer....huge claims will always exist...

i am myself a lil skeptical about 2017 time....i guess 2020 seems to be a more realistic deadline....

as far as petaflop capability is concerned...i guess India is gonna gatecrash the club in 2014 by my best estimates....

as far as starting 4 years ago....well India since its independence has been having "5 year plans"....every 5th year the fund allotment is done for everything....the 11th five year plan ended in 2012....and heres the 12th five year plan in which (officially) the supercomp will be developed

every huge investment is according to the "five year plans"