The bottom line is that the President's FY18 budget proposes to spend $508 million on exascale-related activities. This is a 77 percent increase over the FY17 enacted levels. The intent of this funding is to put the U.S. on track to have a productive exascale system by 2021. Funding is divided between two DOE programs, the Office of Science and the semi-autonomous National Nuclear Security Administration (NNSA). The NNSA request directs $161 million for the Advanced Simulation and Computing (ASC) program and another $22 million to begin construction of the physical infrastructure for the exascale system. The Office of Science (SC) money ($347 million) would go to the Advanced Scientific Computing Research (ASCR) program. (See Tiffany Trader's coverage in HPCwire for a detailed look at the numbers.)
Both the NNSA and SC exascale activities will be the subject of debate as the President's FY18 budget request moves forward in Congress. However, given the cuts that were seen in the rest of the DOE budget, getting to this point could be considered a minor miracle. Getting the increases to the NNSA exascale budget was likely to be relatively easy. President Trump said he was going to increase the federal government budget's emphasis on national security and set aside about $1 billion for the NNSA. Using part of that to add to the ASC program must have been straightforward. That being said, there must have been a tremendous amount of work and planning needed to create the budget justification material.
The Oak Ridge National Laboratory will deploy Summit in 2018, with an estimated 200 petaflops of performance. The Argonne Leadership Computing Facility at Argonne National Laboratory is planning a 180 petaflop supercomputer, Aurora, to be operational by 2019.
(Score: -1, Troll) by Anonymous Coward on Thursday June 01 2017, @03:33AM (7 children)
The government should not be choosing where society's resources go; only The People should be choosing, through their own decisions with their own capital.
If at all, a government should merely be enforcing the agreements that The People make among themselves, according to their own "private" contracts.
(Score: 2) by idiot_king on Thursday June 01 2017, @03:37AM (5 children)
The "People" choosing where the money goes leads to bottled water being more expensive than flavored corn syrup water.
You really think people can choose more effectively than informed individuals? Look at the previous election.
The government keeps a lid on the "People." If that weren't the case, laws would simply not exist.
(Score: 0) by Anonymous Coward on Thursday June 01 2017, @04:14AM
For flavored corn syrup water, I see it is $26 for 100g. BTW, the flavor I chose is beef heart and brain.
http://www.sigmaaldrich.com/catalog/product/sial/53286?lang=en®ion=US [sigmaaldrich.com]
For water, I see it at $10.90 for 100 mL. This is much cheaper.
http://www.sigmaaldrich.com/catalog/product/sigma/w3500?lang=en®ion=US [sigmaaldrich.com]
(Score: -1, Troll) by Anonymous Coward on Thursday June 01 2017, @04:35AM (2 children)
Do you want a president with Parkinson's disease or epilepsy or worse?
Before you bring it up, I must point out that Snopes completely ignores the blue glasses and the miraculous (hour later) pneumonia recovery. (not they have been impartial for years)
Heck, people who cross her tend to end up dead. The latest is a Florida prosecutor, and before that it was Seth Rich. Normally that kind of crap is 3rd-world.
Plus the main qualification for a US president is 100% putting the USA first, and thus there was only one person who met that bare minimum standard, even considering people who didn't get past the primaries.
If you are unaware of the above, then you are not one of the "informed individuals" that you probably fancy yourself as.
(Score: 2) by jmorris on Thursday June 01 2017, @04:49AM
Lets not get too far out with the hyperbole, K? Pretty sure Rand Paul would have put America First. Or at least he would have tried to to the best of his ability and understanding. He wouldn't have beaten HRC and would have been totally hapless if somehow he had won and had to face the Swamp. And of course we have to keep in mind that, like his father, he is batshit crazy. But he isn't Anti-American.
Pretty sure, for all that he was born in Canada and has a Cuban father, Ted Cruz is 100% Pro American and would have done his spergy best. But would also have been devoured by the Evil that is the Clinton Machine and would have been totally outside his competence in dealing with the Lovcraftian Horror that Dwells in the Swamp had he lucked out and survived HRC.
The rest I have my doubts about, except Jeb! who I know for sure can't be trusted.
(Score: 0) by Anonymous Coward on Thursday June 01 2017, @06:03AM
Total nonsense. How does continuing to allow the existence of unconstitutional mass surveillance, the TSA, the drug war, countless unnecessary wars, etc. put the USA first? Or do you not consider the Constituton to be important? To me, any candidate that is fine with violating the Constitution is 100% disqualified instantly, which means Trump, Clinton, Obama, and all their buddies are/were totally unqualified.
Trump is not a good president or even a mediocre president; he is just another authoritarian scumbag. Those who care about liberty do not support Trump.
(Score: 1) by khallow on Thursday June 01 2017, @05:08AM
And what's supposed to be the problem with that?
Who aren't actually so informed? Yes, I do think people can choose more effectively. Let us also note the power of comparative advantage. It is not the best use of our society's resources to have the best and brightest delegated to telling me when I can wipe my ass. They could be doing something more productive like say, developing commercial fusion power. Meanwhile, I could just muddle through my life at little loss to society over the alternative.
(Score: 1) by realDonaldTrump on Thursday June 01 2017, @04:09AM
The people chose me, and now I'm keeping my promises to them. They love my new budget, because it's exascale. The people love it. Paul Ryan loves it. Ted Cruz loves it. Harrisburg loves it. You should have seen the crowd when I went to Harrisburg, lt me tell you. And Bernie. He hasn't said so yet, but I know he'll love my budget. Believe me. Because it's exascale. It's huge. A great budget, part of making America great. A promise kept. Tremendous!
(Score: 3, Insightful) by c0lo on Thursday June 01 2017, @03:37AM
A clickbaity title.
My first reaction: budget in the exadollars range or budget for exaflops computing?
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 3, Funny) by jmorris on Thursday June 01 2017, @04:03AM (2 children)
We wouldn't want there to be a flop gap, would we?
(Score: 2) by c0lo on Thursday June 01 2017, @04:29AM
Of course! We can't afford to miss a flop!!
Every flop counts; from flop to flop towards final victory!!!
(grin)
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 2) by sgleysti on Thursday June 01 2017, @03:43PM
A similar funny response was posted in response to an article about China testing an emdrive in space: "We cannot allow an emdrive gap."
As long as the "we cannot allow a(n) _____ gap" thing is applied to peaceful and useful technologies, I think it's a good thing. It's like sports team rivalries turned toward productive ends.
(Score: 0) by Anonymous Coward on Thursday June 01 2017, @04:41AM (5 children)
Why do we need "exascale" computing? I know they are used for nuke simulation and other research, and we need'em, but what's the point of marketing "exascale" as if it's some kinda meaningful milestone?
(Score: 0) by Anonymous Coward on Thursday June 01 2017, @05:19AM
Man, you must have so many problems understanding human society. Birthdays, anniversaries, rounding numbers up and down, the concept of a decade, century and millennium, megabytes and gigabytes...
Here, let me explain: people like round numbers :)
(Score: 2) by takyon on Thursday June 01 2017, @05:26AM (3 children)
Most supercomputers are multi-user. If a simulation of a desired quality/complexity in some branch of science/engineering would take thousands of hours to run, it won't happen. So increasing computing power by 1 or 2 orders of magnitude, for example 10 petaflops to 1 exaflops, could have a great effect on the kinds of science that can get done. If you look at tasks such as weather modeling, brain simulation, particle simulations, or protein folding, the problems can only be simplified so much. At some point you need to have more parallel processing capability. So I can find it somewhat credible when supercomputing centers brag about what kinds of science 1 petaflops or 1 exaflops can enable.
It seems like major supercomputing centers want to make sure the next supercomputer they deploy is 10x more powerful than the previous. Buying a state-of-the-art 2 petaflops machine and then a state-of-the-art 5 petaflops machine would be wasteful (I specify "state-of-the-art" so as to not include demo machines like the 8.5 petaflops supercomputer "Theia" which is like the test run for a 180 petaflops "Aurora"). So it's natural that there are comparisons to these arbitrary milestones. And remember, the word is "exascale", which you might consider to encompass machines from a broad range of 500 petaflops to 500 exaflops.
Maybe the word is thrown around as marketing to a very specific group of people: Congress. But the national prestige angle doesn't stop at 1 exaflops and countries want to move quickly to 10 exaflops, 100 exaflops... as far as Moore's law will allow. It's not terrible if the U.S. reaches a certain milestone 2 years late; it's only bad if we are say, 5 years late.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 1) by realDonaldTrump on Thursday June 01 2017, @07:23AM (1 child)
Buying a state-of-the-art 2 petaflops machine and then a state-of-the-art 5 petaflops machine would be wasteful
I put mine on Craigslist when I upgrade. I don't keep them long so there's always someone who wants them.
(Score: 2) by Hartree on Saturday June 03 2017, @02:55AM
Nah. Then people get upset that you're dumping e-waste on the third world the second or third time it gets freecycled..
(Score: 2) by Hartree on Saturday June 03 2017, @03:11AM
Back in the early 90s, I worked with a group doing protein folding simulations. We could've easily saturated any computer at that time, but the limitation as you say was budget. We would have bankrupted not only our PI but our whole department. Making computing cheap is just as important as making it fast.
(Score: 2) by Hartree on Saturday June 03 2017, @03:05AM
We're probably not going to get rid of nuclear weapons. And paying a relatively small amount (on government scale) for the next few orders of magnitude in speed so they can keep simulating rather than doing live testing of nukes seems like a pretty good deal to me.
And that's not even counting a lot of the scientific benefits. Faster computers are a godsend for the physics, astronomy, biochemistry, etc. departments. (Assuming the problems are amenable to mass parallelism. That there are problems that aren't gets lost sometimes in the rush towards more processors.)