Obama's New Executive Order Says the US Must Build an Exascale Supercomputer

Jason Koebler writes: President Obama has signed an executive order authorizing a new supercomputing research initiative with the goal of creating the fastest supercomputers ever devised. The National Strategic Computing Initiative, or NSCI, will attempt to build the first ever exascale computer, 30 times faster than today's fastest supercomputer. Motherboard reports: "The initiative will primarily be a partnership between the Department of Energy, Department of Defense, and National Science Foundation, which will be designing supercomputers primarily for use by NASA, the FBI, the National Institutes of Health, the Department of Homeland Security, and NOAA. Each of those agencies will be allowed to provide input during the early stages of the development of these new computers."

Comments

In 30 years, this is our next cell phone.

[GoodNewsJimDotCom]

But can it run Crysis?

Re:In 30 years, this is our next cell phone.

[Jumunquo]

No, but it can figure out you suck at the game.

Re:In 30 years, this is our next cell phone.

[Gravis Zero]

But can it play Crysis?

ftfy. :)

Re:In 30 years, this is our next cell phone.

[davester666]

Yeah, you'll just have to wait for the aliens to arrive. They should have the technology capable of running it at 30 fps on a HD screen.

Re:In 30 years, this is our next cell phone.

[Coren22]

I can, why can't you?

Re:In 30 years, this is our next cell phone.

[davester666]

I'm not an alien.

Re:In 30 years, this is our next cell phone.

[Coren22]

That does explain some things...I always wondered why people give me such odd looks when I walk around in public.

And the NSA?

[Otome]

What would the existence of an exascale supercomputer mean for today's popular encryption/hashing algorithms?

Re:And the NSA?

[Fallen Kell]

A PL3/PL4 nightmare for the people building and running it...

Re:And the NSA?

[quenda]

violate the spirit?

I just mean it enables the US to develop new weapons, e.g. bunker busters, without live testing. Yes, the simulations are that good. I'm not saying this is necessarily a bad thing, especially as the Comprehensive Nuclear-Test-Ban Treaty was never ratified. But the NPT is a problem.

Re:And the NSA?

[DigiShaman]

For that, you would be using custom ASIC hardware, and lots of it.

Re:And the NSA?

[LWATCDR]

Probably none at all. If you want to break today's encryption/hashing algorithms you would probably be using ASICs if not those then FPGAs with GPU compute being your last choice.
Dedicated hardware is the most efficient when you are dealing with a well known standard. For all we know IBM is still in business because it is building NSA ASICs using that 7nm process they showed.

Also time on this beast will be extremely expensive if they use it for any kind of code breaking it will not be for random slashdot users.

Re:And the NSA?

[swillden]

What would the existence of an exascale supercomputer mean for today's popular encryption/hashing algorithms?

Nothing, nothing at all.

Suppose, for example that your exascale computer could do exa-AES-ops... 10^18 AES encryptions per second. It would take that computer 1.7E20 seconds to brute force half of the AES-128 key space. That's 5.4E12 years, to achieve a 50% chance of recovering a single key.

And if that weren't the case, you could always step up to 192 or 256-bit keys. In "Applied Cryptography", in the chapter on key length, Bruce Schneier analyzed thermodynamic limitations on brute force key search. He calculated the amount of energy required for a perfectly efficient computer to merely increment a counter through all of its values. That's not to actually do anything useful like perform an AES operation and a comparison to test a particular key, but merely to count through all possible keys. Such a computer, running at the ambient temperature of the universe, would consume 4.4E-6 ergs to set or clear a single bit. Consuming the entire output of our star for a year, and cycling through the states in an order chosen to minimize bit flips rather than just counting sequentially, would provide enough energy for this computer to count through 2^187. The entire output of the sun for 32 years gets us up to 2^192. To run a perfectly-efficient computer through 2^256 states, you'd need to capture all of the energy from approximately 137 billion supernovae[*]. To brute force a 256-bit key you'd need to not only change your counter to each value, you'd then need to perform an AES operation.

Raw computing power is not and never will be the way to break modern crypto systems[**]. To break them you need to either exploit unknown weaknesses in the algorithms (which means you have to be smarter than the world's academic cryptographers), or exploit defects in the implementation (e.g. side channel attacks) or find other ways to get the keys -- attack the key management. The last option is always the best, though implementation defects are also quite productive. Neither of them benefit significantly from having massive computational resources available.

[*] Schneier didn't take into account reversible computing in his calculation. A cleverly-constructed perfectly-efficient computer could make use of reversible circuits everywhere they can work, and a carefully-constructed algorithm could make use of as much reversibility as possible. With that, it might be feasible to lower the energy requirements significantly, maybe even several orders of magnitude (though that would be tough). We're still talking energy requirements involving the total energy output of many supernovae.

[**] Another possibility is to change the question entirely by creating computers that don't operate sequentially, but instead test all possible answers at once. Quantum computers. Their practical application to the complex messiness of block ciphers is questionable, though the mathematical simplicity of public key encryption is easy to implement on QCs. Assuming we ever manage to build them on the necessary scale. If we do, we can expect an intense new focus on protocols built around symmetric cryptography, I expect.

Re:And the NSA?

[swillden]

For that, you would be using custom ASIC hardware, and lots of it.

No, for that you just laugh at the guy asking you to do it, and look for ways to steal the key, rather than brute forcing it. Even if an ASIC solution gets to way beyond exascale, say to yottascale (10^6 times faster than exascale), you're still looking at on the order of a million years to recover a single 128-bit AES key, on average.

Brute force is not how you attack modern cryptosystems. More detail: http://tech.slashdot.org/comme...

Re:And the NSA?

[swillden]

Probably none at all. If you want to break today's encryption/hashing algorithms you would probably be using ASICs if not those then FPGAs with GPU compute being your last choice.

ASICs, FPGAs and GPUs are all utterly, utterly inadequate to attack today's encryption and hashing algorithms. Unless you have not only tens of billions of dollars but also don't mind waiting millions of years. http://tech.slashdot.org/comme....

Re:And the NSA?

[Khashishi]

Actually, they probably included a few big wrenches to assemble some of the rack systems, so they probably have the tools to break even 1024 bit encryption.

Re:And the NSA?

[swillden]

Actually, they probably included a few big wrenches to assemble some of the rack systems, so they probably have the tools to break even 1024 bit encryption.

When you say "1024-bit encryption" you're talking about RSA, which is a completely different problem. 1024-bit RSA are too small to be used today and should be replaced.

2048-bit RSA keys, however, are roughly equivalent in security against brute force to a 112-bit symmetric key, and will be secure against anyone for quite some time. 3072-bit RSA keys are equivalent to a 128-bit symmetric key. Excascale, even yottascale, computers won't touch them.

But everyone really should be moving away from RSA anyway. ECC is better in virtually every respect. To get 128-bit security (meaning equivalency to 128-bit symmetric key), you only need a 256-bit EC key.

It's for the ...

[CaptainDork]

... NSA data center and stuff.

Re:It's for the ...

[unixisc]

This could bring Intel/HP's Itanium back to life. Make a center w/ 8192 Itanium CPUs, and all the storage needed. Have that located in AK, to make cooling easier.

But...

[Anonymous Coward]

Will it blend?

Re:But...

[Mashiki]

No this version only puree's.

some of the challenges

[necro81]

IEEE Spectrum had an article last year describing some of the challenges we'll need to overcome in order to achieve exascale computing.

Here's another, somewhat pessimistic piece they posted in 2008 - a digest of a DARPA report that went into significant technical detail.

The biggest hurdle is power, and the biggest driver of that isn't the actual computation (i.e., the energy to perform some number of FLOPS), but rather moving that data around (between cores, to/from RAM, across a PCB, and among servers). Other hurdles include how to manage so many cores, ensure they are working (nearly) concurrently, how to handle hardware failures (which will be frequent given the amount of hardware), and writing software that can even make use of such technology in anything approaching optimal fashion.

Not to say its impossible, merely hard given the present state of things and projecting a bit into the future. But as we know, "it is difficult to make predictions, especially about the future." [source?]

Re:some of the challenges

[ILongForDarkness]

Admittedly not as large but I worked on 2000+ node clusters in the early oughts. They way they got "efficiently used" was they were broken up and jobs generally only used a small subset of relatively adjacent nodes. One scientist would use 40 cores on 10 servers sharing a switch, another 100 copies of a serial app on 100 cores etc. Every once and a while, and it was rare, an astrophysicist or whatever would actually use hundreds of cores concurrently for a parallel algorithm. It was by far the minority case though. That said it is ~13 years later and more tools to use parallel exist, and DOE, defense etc have more budget to pay for massive parallel code to be written and even if it is relatively junk if it runs faster in parallel throw 1B of computer at it and say "well our budget is +- 1 steath fighter anyways so ....".

Re:some of the challenges

[LWATCDR]

Sounds like we need higher performance per core. Not all problems are highly parallel, even with those that are you have limits, and now the interconnects are getting to be an issue.

30 Times Faster?

[TechyImmigrant]

For most specific problems thrown at supercomputers, you can go 30 times faster with a custom hardware architecture baked into silicon

To go 30 times fast for general purpose supercomputing, you use the latest silicon (2X) and more chips (15X) and come up with a super new interconnect to make it not suck. This would involve making some chips that support low latency IPC in hardware.

They are free to send me a few billion dollars, I'll get right on it and deliver a 30X faster machine and I'l even use some blue LEDs on the front panel.

Re:30 Times Faster?

[Jumunquo]

The front panel and paint job are the highest margin part of the whole system. You would never give anything there away for free.

Re:30 Times Faster?

[MerlynEmrys67]

You apparently don't know what HPC stands for... Half Price Computing - everyone drives their margins to 0 to win one of these multi million dollar contracts. The only companies that win are the suppliers to the bidders. So lets see in 4 years CPU/GPUs will have 5-6x more compute power, we can throw 5-6x more CPUs at the problem and connect them up with nice 200-400Gbs (5-8x) interconnect. This isn't a research problem - just an interesting engineering problem.
What is probably a research problem is adapting the algorithms to keep all of those 100's of thousand CPU cores busy solving the same problem and keeping them filled with useful work. I am not sure how well MPI will scale to this level

Re:30 Times Faster?

[Jumunquo]

Lol, you mean like IBM with Blue Waters?

Re:30 Times Faster?

[John.Banister]

For most specific problems thrown at supercomputers, you can go 30 times faster with a custom hardware architecture baked into silicon

Perhaps that's what they should do. Make a robotic silicon wafer fabrication facility part of the computer. After being given a task requiring a new architecture, it creates the architecture it needs and augments itself. I'm sure for less than the cost of the F-35 program, a universally tasking self augmenting supercomputer could be made to happen.

Re:30 Times Faster?

[Khashishi]

and it will be useless for everything except one problem

Likely a new gift for the NSA

[mrflash818]

What would the existence of an exascale supercomputer mean for today's popular encryption/hashing algorithms?

Exactly.

My first thought was the new addition will be tasked by the NSA/FiveEyes to break encryption for intercepted communications.

Re: Likely a new gift for the NSA

[Anonymous Coward]

AC is angry at Obama because Obama put AC on a watch list? Which came first? Help me out here.

Re:Likely a new gift for the NSA

[Anonymous Coward]

if the new computer is 30x faster than the fastest one currently deployed and in use.. you've got

1x for weather (noaa),
1x for health (nih),
1x for science (nsf),
1x for nasa, and
1x for energy (doe);

and each of those organizations will be thrilled at having the extra computational power.. that leaves the equivalent of 25 left over for the unconstitutional, illegal, and/or classified shit that they really want it for. the legitimate uses is what they use to sell it and justify its expense, while distracting everyone from what they're really going to use it for (i.e. its true intended purpose).

Re:Likely a new gift for the NSA

[TigerPlish]

Weather guys want this after NSA's done.

We'll take a side of phased-array weather radar to go with that, too.

Re:Likely a new gift for the NSA

[Orp]

Weather guys want this after NSA's done.

I'm a weather guy - running cloud model code on Blue Waters, the fastest petascale machine for research in the U.S. I don't think we've managed to get any weather code run much more than 1 PF sustained - if even that. So it's not like you can compile WRF and run it with 10 million MPI ranks and call it a day. Ensembles? Well that's another story.

Exascale machines are going to have to be a lot different than petascale machines (which aren't all that different topologically than terascale machines) in order to be useful to scientists and in order to no require their own nuclear power plant to run. And I don't think we know what that topology will look like yet. A thousand cores per node? That should be fun; sounds like a GPU. Regardless, legacy weather code will need to be rewritten or more likely new models will need to be written from scratch in order to do more intelligent multithreading as opposed to mostly-MPI which is what we have today.

When asked at the Blue Waters Symposium this May to prognosticate on the future coding paradigm for exascale machines, Steven Scott (Senior VP and CTO of Cray) said we'll probably still be using MPI + OpenMP. If that's the case we're gonna have to be a hell of a lot more creative with OpenMP.

Re:Likely a new gift for the NSA

[Anonymous Coward]

What horseshit. Weather modelling takes a very, very different topology than the NSA's toys.

And what mission does the FBI have? I got all the rest (DHS has some WMD modelling), but I fail to see a justifiable reason for the FBI to have a supercomputer.

Re:Likely a new gift for the NSA

[youngatheart]

I think you underestimate the complexity of modern encryption and hashing algorithms.

Re: Likely a new gift for the NSA

[lucm]

Hopefully you don't think of yourself as more tolerant than him.

Re:Likely a new gift for the NSA

[rtb61]

Artificial intelligence to flood the internet with pro American Corporate propaganda upon a massive scale, unfortunately that is not a joke but a serious intent.

Re:Likely a new gift for the NSA

[aaaaaaargh!]

How would you run secret programs on a computer shared with NOAA and NSF? The NSA don't need it, they have their own supercomputers. Even their budget it secret.

Re: Likely a new gift for the NSA

[thaylin]

Except the OP was not just unhappy with Obama, he went out of his way to use logical fallacies.

Re:Likely a new gift for the NSA

[Skapare]

With OpenStack.

Re:Likely a new gift for the NSA

[dj245]

Weather guys want this after NSA's done.

I'm a weather guy - running cloud model code on Blue Waters, the fastest petascale machine for research in the U.S. I don't think we've managed to get any weather code run much more than 1 PF sustained - if even that. So it's not like you can compile WRF and run it with 10 million MPI ranks and call it a day. Ensembles? Well that's another story.

Exascale machines are going to have to be a lot different than petascale machines (which aren't all that different topologically than terascale machines) in order to be useful to scientists and in order to no require their own nuclear power plant to run. And I don't think we know what that topology will look like yet. A thousand cores per node? That should be fun; sounds like a GPU. Regardless, legacy weather code will need to be rewritten or more likely new models will need to be written from scratch in order to do more intelligent multithreading as opposed to mostly-MPI which is what we have today.

When asked at the Blue Waters Symposium this May to prognosticate on the future coding paradigm for exascale machines, Steven Scott (Senior VP and CTO of Cray) said we'll probably still be using MPI + OpenMP. If that's the case we're gonna have to be a hell of a lot more creative with OpenMP.

I'm not a weather guy, but my understanding is that a somewhat fixed weather model (set of calculations) is used to do a kind of finite-element analysis on small areas. With better computing and better radars, smaller and smaller areas can be calculated, which results in more accuracy.

With more computing power, could you not vary the parameters or constants used in the weather model, then run the finite-element analysis over the entire weather area again? You could be running hundreds or thousands of slightly different weather models, then apply some processing to figure out which is most likely- either by averaging together the 50% most similar outcomes, or by some other method. I don't think you could peak out a supercomputer with that method if you kept adding more parameter variations, although you may get to the point where adding more parameter variations doesn't improve accuracy.

Maybe that's an incorrect understanding, but we're getting closer to the point where we can calculate all possible outcomes simultaneously. I wouldn't have expected this to be the case with weather but computing has come a long way in the last 20 years.

Re:Likely a new gift for the NSA

[Orp]

You can say "one of" but you can't say "the fastest" petascale machines my friend

http://www.hpcwire.com/2012/11...

I should have added "on a college campus".

My main point is, just throwing more cores at "mostly MPI" weather models is not sustainable. We are going to need to be much smarter about how we parallelize.

Re:Likely a new gift for the NSA

[Orp]

One of the biggest problems of the current large scale HPC machines is users (like you but maybe not you specifically) are typically scientists/analysts who write software that does not scale well. There either needs to be better frameworks for you to work within that handle all the grunt work of doing efficient parallelization and message passing or every atmospheric physicist needs to be teamed with a computer scientist and a software engineer.

Absolutely agree 100%!

Re:Likely a new gift for the NSA

[Orp]

You are basically describing ensemble forecasting, which is very powerful for providing statistically meaning forecasts where you can intelligently talk about the uncertainty of the forecast, something single deterministic forecasts cannot do.

In my research, I'm doing single deterministic forecasts to study what happens with tornadoes in supercell thunderstorms, where I am cranking up the resolution to capture flow that is otherwise unresolved. I get one version of a particular storm, which is good for studying certain aspects of storms, but not good at being able to generalize (that takes lots of simulations).

Both big deterministic simulations and ensembles have their place. Of course, today's big simulation can be the resolution of tomorrow's ensembles! Right now, you can do lots of good science with ensembles. Operationally (weather forecasting) this is basically the new paradigm, although forecasters are slow to change from just looking at the single deterministic GFS and NAM forecasts. The ensemble approach, once we start running hundreds of forecasts at higher resolution that we do today, will transform our forecasting accuracy (and precision). However it will be limited to the amount of good observational data we can feed the models (otherwise GIGO). This is where remote sensing comes in. GOES-R will be a big help.

It will indeed take people from atmospheric science, computer engineering, software engineering, etc. working together to best exploit exascale machines. NCSA understand this and that's what makes it (and other similar organizations) great.

Just think of the carbon footprint of this...

[zkiwi34]

Mind you, think of the movies it could do in a bazillionK resolution.

Encryption

[koan]

How long would it last against that?

Re:Encryption

[Anonymous Coward]

If we've done it right, quite a long time ?.

RSA for example the difficulty goes up exponentially with bits.

So 4k RSA is 2^2048 times harder than 2k RSA.

2^2048 is a number a LOT bigger than 30.

Re:Encryption

[Anonymous Coward]

Is it bigger than 40?

Re:Encryption

[ciaran2014]

Is it bigger than 40?

Depends on the value of ^

Re:Encryption

[rthille]

Facepalm

Re: Gotta love these executive orders

[Anonymous Coward]

And random person freaks about because President exercises his lawful authority to tell agencies and departments under his jurisdiction to cooperate and present a plan for creating a supercomputer.

Here is a hint:

Sec. 7. General Provisions. (a) Nothing in this order shall be construed to impair or otherwise affect:

the authority granted by law to an executive department, agency, or the head thereof; or
the functions of the Director of OMB relating to budgetary, administrative, or legislative proposals.
(b) This order shall be implemented consistent with applicable law and subject to the availability of appropriations.

(c) This order is not intended to, and does not, create any right or benefit, substantive or procedural, enforceable at law or in equity by any party against the United States, its departments, agencies, or entities, its officers, employees, or agents, or any other person.

It is like nobody knows how the government operates any more, but if Obama does it, they're opposed, damn opposed.

Re:nuclear weapons

[Jumunquo]

... that we never use...

Re:Why not just call it 1 exaflop computer?

[PaulBu]

Because 0.1 exaflops is still "exascale", but not "exaflops"... :)

Paul B.

Classified Data

[Etherwalk]

What would the existence of an exascale supercomputer mean for today's popular encryption/hashing algorithms?

Exactly.

My first thought was the new addition will be tasked by the NSA/FiveEyes to break encryption for intercepted communications.

Why are you assuming they don't already have one doing that, and this is just a public version?

There is a lot of highly secured government data infrastructure out there that I hear about even though not inquiring. The cable in Virginia that gets cut by a backhoe accidentally and guys in a black van show up ten minutes later. The contract for a government data center inside a faraday cage. The government likely already has much more computing power available than we know about.

Re:Classified Data

[MobSwatter]

"First rule in government spending... Why build one when you can have two for twice the price".

Looks like Skynet or nothing folks. But with two we're going to have Jade Helm and martial law bankster employee's with two personalities, the dominant one will likely be the one that screws you out of the most money.

Re:Classified Data

[ILongForDarkness]

I agree. The very fact we'll know where this datacentre is probably means it will be being used for relatively lower security stuff. The exascale supercomputer for actually analyzing the NSA intercepts probably already exists.

Re:Classified Data

[AmiMoJo]

We can be fairly sure that the NSA has some serial dedicated hardware for cracking common encryption systems like AES. They will still be reliant on things like dictionary attacks because brute-forcing the entire keyspace is impractical (unless they have quantum computers).

How should we react to that? Well, obviously we need a good password that can resist dictionary attacks. Beyond that, unless you are a big enough perceived threat to warrant time on an expensive computer you probably don't have to worry too much. They certainly won't be using it to help out the FBI, risking its existence coming to light.

Re:Classified Data

[Coren22]

Serial? Must be pretty fast to do operations of that kind in serial.

Re:Classified Data

[Etherwalk]

We can be fairly sure that the NSA has some serial dedicated hardware for cracking common encryption systems like AES. They will still be reliant on things like dictionary attacks because brute-forcing the entire keyspace is impractical (unless they have quantum computers).

How should we react to that? Well, obviously we need a good password that can resist dictionary attacks. Beyond that, unless you are a big enough perceived threat to warrant time on an expensive computer you probably don't have to worry too much. They certainly won't be using it to help out the FBI, risking its existence coming to light.

Maybe. Based on the documentaries that have been made, it's pretty clear that the NSA used their phone-metadata-recording to help the FBI locate the Boston Bomber, despite the risk that it would become public. (Which is did shortly thereafter but for other reasons--i.e. Snowden).

The FBI does domestic counterterror. The NSA is the big bad in terms of not seeing the inherent bad and threat to democracy in snooping on everyone's communications, sure, but they're still trying to be good guys and so they'll share information sometimes when they see a good result from doing so.

Inflammatory Remark Warning

[Jack Griffin]

So can someone tell me, is better or worse than than banning stem cell research?

Re:Inflammatory Remark Warning

[digsbo]

I guess it depends on whether you think the Federal Government should be allocating tax dollars for things other than basic defense and infrastructure.

Re:Inflammatory Remark Warning

[Jack Griffin]

Stem Cell Research and Computational power are all defence. The stronger and more capable you are as a nation, the more effective your self defence.

Re:Inflammatory Remark Warning

[digsbo]

That argument can be stretched to absurdity, though. And already has been.

Re:Inflammatory Remark Warning

[Jack Griffin]

Yes but that's the benefit of being able to think about and discuss things as rational humans, we can work out which concepts have value and which do not.
IMO both advanced medical research and computational power are must haves in any powerful democracy.

Re:Inflammatory Remark Warning

[digsbo]

"Powerful democracy" is an interesting phrase, and there's no reason to believe either computation or medicine can't be provided by the market. Wouldn't you simply want the benefits of computing power and advanced medicine?

Re:Inflammatory Remark Warning

[Jack Griffin]

there's no reason to believe either computation or medicine can't be provided by the market.

Except for the fact that the most free market medical system in the western world is also one of the worst performing. Apart from that, yeah it's all good...

Exascale? We don't need that.

[the_humeister]

640 petaflops ought to be enough for anybody.

Department of Homeland Security

[Anonymous Coward]

Who in their right minds would let these people near a computer? Please. Let them go back to what they excel at: stealing cameras out of our luggage and groping underage genitalia.

Re:Department of Homeland Security

[cdrudge]

You do realize that the TSA is only about 1/4 of DHS by number of employees and 12% of the budget, right? I think you're really selling short the amount of damage they excel at if you only go with stealing cameras and groping underage genitalia.

Re:Department of Homeland Security

[Hodr]

Seems much more likely that this would be used by CBP and the Coast Guard (both DHS).

Might take a decent amount of horsepower to constantly search a database for every tracked vessel and run an analysis on each to determine when they do something that is out of the ordinary, then compare those results to patterns that predict some form of unwanted behavior (trafficking, illegal fishing, hijacking, lost at sea, etc.)

easy b/c avg time from order to delivery 4.5 years

[raymorris]

Those issues will be resolved by a side effect of this being a government order. According to the GAO, on average it takes 4 1/2 years from the time the government orders a computer until it's installed. Right now, multiple government agencies have been told to start thinking about a plan. In two years (2017), each agency will have their plan and they'll start working to to resolve the differences between agencies. In another year (2018), they'll put out some RFPs. Those will go through the federal procurement process and the order will be placed about two years later (2020). That's when the 4 1/2 year average clock starts, so expect installation around first quarter 2025.

The goal is that it should be 30 times faster than TODAY'S computers.
And be operational in ten years. They can pretty much just order a Nexus 47, or an HP Proliant gen 12.

Skynet

[Bodhammer]

John Connor: "By the time Skynet became self-aware it had spread into millions of computer servers across the planet. Ordinary computers in office buildings, dorm rooms; everywhere. It was software; in cyberspace. There was no system core; it could not be shutdown. The attack began at 6:18 PM, just as he said it would. Judgment Day, the day the human race was almost destroyed by the weapons they'd built to protect themselves. I should have realized it was never our destiny to stop Judgment Day, it was merely to survive it, together. The Terminator knew; he tried to tell us, but I didn't want to hear it. Maybe the future has been written. I don't know; all I know is what the Terminator taught me; never stop fighting. And I never will. The battle has just begun."

Name of the Computer Project

[Fire_Wraith]

Hm, something like this needs a good, catchy name, that also makes for a good acronym.
How about...

Strategic
Kinetic
Yankee
Neural
Exaflop
Terminal

Re:Name of the Computer Project

[Overzeetop]

Did I catch a "niner" in there?

Imagine a

[Rinikusu]

beowulf cluster of these...

Re:Imagine a

[Pfhorrest]

A five digit UID making a Beowolf Cluster joke? What decade is this!?

Re:Robocop anyone?

[roc97007]

I think "galactically stupid politicians" is my new favorite term.

Re:Linux, Linux, Linux, Linux

[roc97007]

...unless they skip "11"....

So...

[pem]

I guess Obama's miffed he missed out on the early days of cheap bitcoin mining.

Fix the economy

[ChadSmith4920]

Mine Bit Coins

Re: Gotta love these executive orders

[Anonymous Coward]

Just ignore the parent poster. When someone like them disagrees with an executive order, they bitch, moan and make snide remarks about it even if time proves it to be a good thing. When they agree with an executive order, they bitch, moan and make snide remarks about people who disagree with it even if there aren't any around.

In short, they will always bitch, moan and make snide remarks; and are generally unpleasant people. I attribute my happiness in part to just staying away from them :)

The next executive order

[ChrisMaple]

Synthesize the unicorn genome, to provide fuel for transportation and buy the votes of little girls.

Do you want Skynet?

[ArcadeMan]

Because that's how you get Skynet.

Re:Do you want Skynet?

[roger10-4]

AI is how you get Skynet; not HPC. HPC and AI are different problems. While it would be reasonable to assume an AI could leverage the compute power from an exascale platform, it's not necessarily a requirement in achieving it.

I don't want to be serious but...

[ExXter]

... doesn't that sound like Skynet Alpha?

This order is worthless without funding

[acoustix]

He can attempt to mandate all he wants. Congress approves the budgets. And since we all know how well Obama has been submitting his budgets....

Re:This order is worthless without funding

[acoustix]

If you even had a basic idea of how the Constitution works, the President's budget is basically for show. Congress is entirely responsible.

https://www.youtube.com/watch?v=KIbkoop4AYE

Sorry, but you're wrong.

Congress generally begins its budget process once the President submits his budget. The President is required by law to submit a comprehensive federal budget on or before the first Monday in February (31 U.S.C. 1105(a))

Re:This order is worthless without funding

[DigiShaman]

Julius Caesar says, Julius Caesar gets!

Re:This order is worthless without funding

[WindBourne]

yeah, so.
When has the neo-cons/tea* listened to any budget by Obama? Not a once.
They discard everything and simply run their own. Hell, they do not even listen to past GOPs BEGGING for the house/senate to raise taxes on fuel to bring our roads back to levels.
At this time, all they listen to, are the billionaires, along with Chinese gov.. Hell, this group has been working to kill America's new private space, by giving MORE MONEY to Putin.

Re:This order is worthless without funding

[Straif]

The neo-cons/tea partiers had nothing to do with Obama's budgeting woes. Between the House and Senate his proposed budget has only been able to get about 3 positive votes in the last few years. Most years the Senate under Reid simply refused to bring it to the floor for a vote it was so laughable. Only Republican pressure managed to get it to the floor where it managed to get 1 "yea" vote in 3 years.

Only the House actually bothered to do their duty and propose a budget each year. The Senate, once again under Reid;s leadership, generally just ignored all budgetary requirements and simple kept passing short term continuing resolutions which lead to the almost annual face off and threatened government shutdown.

The way the process is suppose to work is the President proposes a budget (usually after some discussion with leaders from both parties which was not going to happen under Obama since his own party members claim he never talks to them), both the House and Senate consider the Presidents proposals and write their own budgets (this can be anything from a direct copy of the Presidents or completely fresh). Then after both house pass a budget they two version go into reconciliation to see if a common ground can be found. While not legally binding, this finalized budget then gives way for appropriation bills to be passed in both houses to fund each budget concern.

For most of Obama's 2 terms things have not worked out that way. The presidents proposals have been seen as a joke by both parties in both houses, the Senate is fine with zero planning and the House passes their own budgets that had no real legal standing and since they know that, can effectively just be a Republican wish list with little or no Dem input.

Re:aaaaaaannd....

[Ol Olsoc]

And democrats have a hard on. Yes President Obama can create anything through executive action.

Just like all the others.

https://en.wikipedia.org/wiki/...

Or do you just want REpublicans to have that?

Rage on!

Re:Dictator dictating is a Dictator

[jeffmflanagan]

Never forget something that exists only in your fevered imagination? What's the point of that when you don't post your bizarre belief under your own name so we know to avoid you?

Windows versions

[rossdee]

Given that even numbers suck, I am sure they will be skipping odd numbers from now on.

I'll stick with Win 7

Re:Windows versions

[roc97007]

Given that even numbers suck, I am sure they will be skipping odd numbers from now on.

I'll stick with Win 7

As will I. They'll need to pry Win 7 from my cold dead fingers. I read a few reviews on 10 yesterday, and the general consensus is that it's almost as good as 7, (better in some parts, worse in others) if you replace that hybrid start menu thingy with Classic Shell and get used to where they've moved things. Yeah. No.

Re:Windows versions

[Valdier]

Given that even numbers suck, I am sure they will be skipping odd numbers from now on.

I'll stick with Win 7

As will I. They'll need to pry Win 7 from my cold dead fingers. I read a few reviews on 10 yesterday, and the general consensus is that it's almost as good as 7, (better in some parts, worse in others) if you replace that hybrid start menu thingy with Classic Shell and get used to where they've moved things. Yeah. No.

Well, seeing how its an extension of Windows 7/8 and there is nothing missing that is in 7 really. Although if new interfaces bother you, it could be a reason to stick with Win3.1

You need to read reviews from professionals, not friends on Facebook ;)

Re:Windows versions

[roc97007]

Given that even numbers suck, I am sure they will be skipping odd numbers from now on.

I'll stick with Win 7

As will I. They'll need to pry Win 7 from my cold dead fingers. I read a few reviews on 10 yesterday, and the general consensus is that it's almost as good as 7, (better in some parts, worse in others) if you replace that hybrid start menu thingy with Classic Shell and get used to where they've moved things. Yeah. No.

Well, seeing how its an extension of Windows 7/8 and there is nothing missing that is in 7 really. Although if new interfaces bother you, it could be a reason to stick with Win3.1

You need to read reviews from professionals, not friends on Facebook ;)

See, this is the real problem. Back in the 3.1 days, we were on the steep end of the curve, and there were things that really needed to be added, changed and fixed. Now we're up on the flat end of the curve, and there really isn't a lot that needs to be improved, assuming that we're still using a keyboard and a mouse and the peripherals haven't changed substantially.

So people went to 95 because the ideas (not necessarily the implementation) really were needed, and to 98SE because 95 kinda sucked, and 2000 because it was stable, and xp for things that 2000 didn't support, and 7 because its 64 bit version was a lot more mature and stable than xp's 64 bit version. There's no *reason* to go to 10, and so no reason for me to muck with live tiles on the start menu and figure out which admin applets are still in control panel and which have been moved elsewhere. It's just not necessary.

We need to remember, Windows is not the application. The applications are the applications. Windows just runs them.

Obligatory #31

[Tablizer]

Imagine a Beowulf Cluster of pork

Re:Just in time for Windows 10

[PPH]

since no other OS can use up all the resources.

FTFY.

It's sad.

[Lord Kano]

The first thought I had was that I hope they don't do it because it'll only be used to invade our privacy.

LK

Re:Linux, Linux, Linux, Linux

[fredgiblet]

They are sticking with 10 forever, like Apple.

The GOP ...

[PPH]

... will throw a wrench in the works.

Buried in some farm bill, there will be a requirement to port systemd to this.

Re:Linux, Linux, Linux, Linux

[lucm]

Don't worry, they are planning to use Oracle Linux. They are currently using the 2nd most powerful computer in the world to calculate how much the license will cost.

Deja vu

[Triakter]

"Each of those agencies will be allowed to provide input during the early stages of the development of these new computers."

Isn't that how we got the $388 Billion broke-dick F-35 Joint Strike Fighter?

Comment removed

[account_deleted]

Comment removed based on user account deletion

Comment removed

[account_deleted]

Comment removed based on user account deletion

and how much will be done in India and China?

[WindBourne]

Seriously, building the computer is worthless, unless it comes 100% from America, or at least the west.
If the parts come from China, then it will make it trivial for China to simply build their own CPLA computer for weapons modeling.

good news?

[prof_robinson]

Oh please...we all know that this is going to be used to spy on us. Why are we cheering?

Re:nuclear weapons

[Dutch Gun]

I'm actually okay with us not using our nukes.

Exascale machines are for scientific computing

[gentryx]

These Peta/Exascale supercomputers are build for computer simulations (climate change, nuclear weapons stewardship, computational drug design, etc.), not for breaking encryption. That's also one reason no one is using them to mine Bitcoins: they're just not efficient at that job. To compute lots of hashes, dedicated hardware designs (read: ASICS) far outpace "general purpose" supercomputers.

Capacity vs. capability

[gentryx]

So, what you describe is essentially the difference between capacity and capability machines. The national labs have both, as there are use cases for both. But the flagship machines, e.g. Titan at the Oak Ridge Leadership Computing Facility (OLCF), are always capability machines -- built to run full system jobs, jobs that scale tens of or hundreds of thousands of nodes.

Re:Capacity vs. capability

[ILongForDarkness]

Thanks for the explanation.

They already co-design the hard-/software

[gentryx]

Basically, the procurement process for supercomputers is like this: the buyer (e.g. a DOE lab) will ready a portfolio of apps (mostly simulation codes) with a specified target performance. Vendors then bid for how "little" money they'll be able to meet that target performance. And of course the vendors will use the most (cost/power) efficient hardware they can get.

The reason why we're no longer seeing custom built CPUs in the supercomputing arena, but rather COTS chips or just slightly modified versions, is that chip design has become so exceedingly expensive and that the supercomputer market is marginalized by today's mainstream market.

Also, the simulation codes running on these machines generally far outlive most supercomputers. The stereotypical supercomputer simulation code is a Fortran program written 20 years ago, which received constant maintenance in the past years, but no serious rewrite is viable (costs exceed price of hardware). So vendors will look for low-effort ways of tuning these codes for their proposed designs. Sticking with general purpose CPUs is in most cases the most cost efficient way.

Re:Dictator dictating is a Dictator

[x0ra]

however, if you look year-to-date, it's still making gains.

Re:but why?

[Shortguy881]

Do you really have to ask that question? The biggest limiting factors to large scale surveillance are technical limitations. This is one step closer to the designed police state. God I sound paranoid. Unfortunately, I'm also right.

FBI? Hmmm.....

[Dr. Blue]

I see people speculating above about the government using this to break crypto, but that's really not a huge concern. If people use good keys that require brute force searching, even the smallest AES key size would take over a billion millenia to break at 10^18 ops/second (even assuming you could test a key on one "op"). And for people who use bad keys, you don't need exascale computing to break them.

So what could the FBI use something like this for? What about analysis of massive public and not-so-public data, like data mining Internet postings, email/phone records, ... Better not post something with the wrong combination of words, or someone might come knocking on your door.

Re:aaaaaaannd....

[Ol Olsoc]

Wait, are you a Republicrat angry because you think he's a Democan, or a Democan angry because you think he's a Republicrat?

Who's angry? I'm just being an agitation engineer.

p.s. to answer your question, I'm a pragmatic. I believe in what actually works, as opposed to ideology. Wildly reviled if the liberal or conservative can expand their mind enough to even acknowledge my existence, I'm the turd in the punchbowl of politics.

Re:Dictator dictating is a Dictator

[Austerity Empowers]

Now if you want to hate on Obama, you could argue that this supercomputer will be designed by indentured servants from India, using components made in Malaysia, and assembled in China. And it will likely be true.

But, you can just call him names too, that's good.

Re:Linux, Linux, Linux, Linux

[Khashishi]

AFAIK all supercomputers use Linux

Re:Quantum computing

[outlander]

The folks at DWave would probably say no:

http://www.dwavesys.com/

They've been working on it for a while. :)