I'm an HPC professional, and do not see much value in these "hero" machines. Yes, you can go on all you want about the march of progress and tier-1 and grand challenges, but you're just reiterating an unquestioned manifest destiny-based view of history. Why do we need an Exaflop machine? is it because some particular set of applications need it? where is the threshold for those applications where the compute facility will be fast enough to achieve some breakthrough?

it's hard to find areas that are primarily limited by compute facilities. for instance, genetics/proteomics/metabilomics/whatever are *not* compute-limited, especially at the high end. they're laboratory-limited, the same way weather simulations are good and getting better, but not past the quality of their input data.

we need more compute in general, but not necessarily in one machine. a single exaflop machine will cost much more than a thousand petaflop machines. letting a thousand flowers bloom is much prettier than one excruciatingly beautiful flower...

and no, hero machines do not provide an efficient way to improve the tech of lesser or later machines. they have to be justified by their own need.

nah. providing wider and faster memory will help even purely CPU codes, even those that are often quite cache-friendly. the main issue is that people want to do more GPUish stuff - it's not enough to serially recalculate your excel spreadsheet. you want to run 10k MC sims driven from that spreadsheet, and that's a GPU-like load.

but really it's not up to anyone to choose. add-in GPU cards are dying fast, and CPUs almost all have GPUs. so this is really about treating APUs honestly, rather than trying to pretend they can survive on old-fashioned CPU memory interfaces.

the current micro-USB connector kinda sucks. if we're going O(100x) more watts, maybe we should take the opportunity to do a better connector, too.

symmetric would be nice, and less prone to jamming, misalignment and torquing.

If a potential user can't manage to install the Linux of their choice onto pretty much any laptop,
they're going to be pretty disappointed trying to actually *run* Linux, even if it's preinstalled.

I've had no problems running Fedora on my Samsung UB.

RMS and like-minded ideologs never really understood that most projects that adopted GPL (like Linux or even GCC) did so for expediency, not belief.

If you can't operate your package manager well enough to install just what you need, you shouldn't be admining anything, VM, or not.

Manhattan or Apollo projects were successful primarily because they had such a clear focus.

Sequencing the human genome was just a way to push development of techniques: we didn't learn that much from the primary product. especially since it's become clear that expression is far more interesting/relevant than just a straight read of sequences. and even that is arguably incomplete without better proteomics.

neuroscience is not at any clearly defined threshold where we can see what's needed to get to a state of much higher understanding. throwing money at the problem will just exercise our inability to pick winners. (and if there's anything fundamental we know about funding, it's that we, especially governments, simply cannot pick winners. this is why broad funding of basic science is essential: we can't know which directions will pay off.)

sometimes, such big-spending programs are just trying to stretch the normal timeline of a technology. is it actually important to spend $1e8 in order to bring about some technical advance by a year? maybe it's 5 years, maybe it's 6 months. in neuroscience, no one even talks about the unknown unknowns, let alone the known unknowns ;)

sure, you can pound or melt, but for my money, sandpapering the platters by hand is the best way to pay homage to your dearly departed bits.

I suppose someone still makes buggy whips, so why not Java version N+1? at least some fraction of this shrinking niche will fail to recognize that programming "investment" is a sunk cost...

the purpose of google glass is obviously creepshots and/or the virtual naked filter. how does wrist-mounting help? maybe the pulse sensor at your wrist can trigger image/video capturs whenever your pulse is elevated? what could ever go wrong with that!?!

pulse and galvanic skin response-driven advertising, where have you been all my internet!

do you really think tax cheating costs less than fixing a power plant or two? why would it matter if they were sponsored by Coke or the Illuminati?

the spectre of foreign hackers taking out infrastructure is asinine: we are responsible for the vulnerability of our systems, just as you are responsible when your system gets hacked...

if a country can kill hackers, shouldn't it be able to call in the drones against tax cheats, dishonest bankers, publishers of unflattering new articles, jokes which insult the dignity of the nation...

it's up to us.

we're the ones who will provide the protocols that would permit the sorts of activities mentioned here to take place in a non-proprietary manner. sure, companies like microsoft seek to dominate their markets, and view lock-in one of the available tools. that's because we let them. we as a society have set up companies to be driven entirely by profit, and have not arranged our legal system to distinguish between proprietary and open systems.

look at tcp/ip, the single most successful open standard in the universe. it didn't just spring fully formed and without peers - there was lots of competition. it won because a few of the companies (and educational institutions and even government) found ways to make it into a world-scale protocol. companies get it if you say "interop is a non-negotiable precondition to purchase". government rightly gets involved not only as significant sales targets themselves, but also when they say (or should), that any utility-type monopolies granted must conform to non-proprietary standards.

imagine if mobile data service was non-proprietary: your phone simply negotiated a 5 minute service contract with the set of carriers it could detect at the moment, wherever you happen to be. (voice and text would simply layer over data, of course.) yes, that sort of thing is obvious to any techie as The Right Way, but it's our fault that the public has gone along the proprietary route: we need to speak up.

business tries to get away with whatever it can - that's just economic darwinism. we just need to set the rules.