What do we use supercomputers for?
Weather forecast
Eroge.
>>7730918
Calculations that are incredibly massive, which can range from realistic graphics simulations (Read the Wiki article over the topic Fast Inverse Square Root), to weather forecasts, to population analysis and pattern recognition.
Dwarf Fortress
>>7730918
bitcoins
Pong.
>>7730931
>run DF on supercomputer
>realize it cant run on more than 1 core
>fps death in minutes
>>7730918
Calculating triple integrals.
>buy supercomputer
>put in my prime finding function, which is just incriminating by -1 and checking if mod the number == 0 until 0
>run it on randomly generated big numbers indefinitely
Inhumane
>>7730918
Downloading porn.
expensive CFD, FEM, theor physics
>>7730918
Encoding pornography.
Making lifelike 3d renders of nude women engaged in sexual acts.
>>7730918
I know neuroscience uses them for large scale neural network modeling.
finally playing cs with const fps ;)
I used to run large scale molecular dynamics simulations on our supercomputer, and also on the cluster. I ran at low priority, and was by far the biggest user in terms of cycles... put us from 5% usage to 98%.
Side note: it was funny how quickly upgrades happened. When I started in the summer after third year, a run took 2 years at a resolution of 1000. (I know, the number doesn't mean anything, but it's for scale.) By the time I finished undergrad and then my master's degree, a full run took 6 weeks at a scale of 3000.
The real-time processes were mostly from the fluid dynamics group, someone doing airshed and weather modelling, and a group doing population dynamics (some sort of insect spreading). They'd take a hundred processors or so for an hour at a time. I saw a few runs from a computational algebra student, but he never really got into it... I think he dropped out or something. The NN guy didn't really need the processing power, and dealing with the batch system wasn't worth the effort.
Browsing dank memes
![facebook-one_2588746k[1].jpg](https://i.imgur.com/IHYmPAC.jpg "facebook-one_2588746k[1].jpg")
>>7730918
We as in humanity? Currently? Mostly facebook.
>>7731103
This is the kind of stuff I wish I studied more. I'd go to grad school for this kind of thing but I don't know if I'm smart enough. Last semester my QM prof had shown the class a simulation her group did wherein you can see the solvent molecules around a molecule which is in a quantum superposition "split" into two different states.
Compiling profiles of citizens based on facebook and twitter posts.
>>7731130
Just posting to mention, I was saying this long, long before it was cool. And everyone called me a paranoid schizophrenic moron, and told me to get help for my delusions.
As usual, I was right. Wish I could kick you cocksuckers around a bit, and watch you drag your broken body as you desperately try to play catch up just to barely get to where I started.
>>7731135
Does pretending to be retarded lasting effects?
Like holding a silly face too long and becoming stuck with it. Can shitposting cause brain damage?
>>7731143
I wouldn't call it outright damage, but it does spur maladaptive changes in the mind. You become predisposed to latch onto and easily generate responses based on criteria and signals that aren't necessarily useful.
Over time it becomes easier and more natural, it becomes a way of thinking you get better at. The path of least resistance can be changed to be suboptimal relative to variable environmental and personal demands.
Going to physically harm you, btw.
Shitposting to the extreme.
>>7731103
What year did you start? By resolution you mean number of particles? MD person here, I've worked on Folding@Home and recently XSEDE (on COMET).
>see thread
>check wiki about supercomputers
>apparently supposed to have exaflops soon thanks to graphics cards
It seems like vidya finally contributed to the world.
>>7730918
simulating the blast of a nuclear weapon
>>7730918
To compute things
I'm processing satellite data. I don't really need a very good computer (it helps obv.), but I need a shitload of them, since there are millions of measurements I need to process.
Shit's getting cumbersome. Big data my arse. Satellites dumping a gorllion measurements per day, how the fudge can we ever keep up?
>>7730918
for creating rare pepes
My plebeian mind is imagining super computer as a super huge Samsung CRT in some basement somewhere. And I cant stop thinking it as so.
>>7731501
Maybe you don't need to process all the data. Maybe you can process half of the data and get 95% of what you need.
>>7731550
We obviously do that already. However, people are constantly asking about better coverage, better resolution, longer time series.
In a way, it keeps me employed, since I'm working on methods to speed up the process, but I don't know where we'll be in 15 years down the line if data throughput increases at the current rate. At some point, only 10 places in this world will be able to actually do the processing because they have 100,000+ cores in their cluster.
>>7730918
To calculate how much your mother weighs
>>7730918
To compute shit
>>7730918
Faster than Alienware for sure.
What if we're living in the matrix, and all those super computers are the machines that our world is running on. Our entire universe, inside a box, inside our universe.
>>7731592
>assume she is looking at someone else
>awkwardly hold her gaze
>try to talk but mouth is too dry
>ironic because face is now wet with sweat
>stammer awkwardly through a sentence that doesnt make sense
>she begins to look confused
>palms shaking, knees weak, put hands in pocket to stop the shaking
>pockets are warm and slippery, spaghetti starts falling out
>cant stop, soon she is covered with spaghetti, entombed in marinara sauce
Molecular dynamics
If we just keep throwing RAM at it will it become intelligent at some point
>>7731592
yeah no they arent powerfull enough
but this little bugger tho could theoreticaly unlock all the secrets of the universe
but its not fully operational yet
behold the first quantum computer
>>7731661
a million times more powerfull then any super computer ever created
a whole room filled with super computers couldent even compare to one of these
Hello Dave
>>7731672
Depends on the size of the room, I could make a very big room
>>7731704
Luke?
>>7731104
Super browsing super dank super memes you mean
>>7731762
Who even are you?
>>7730918
To run Witcher 2 at more than 15 frames per second.
>>7730918
mostly spying but also cracking
>>7731129
>I don't know if I'm smart enough
Nah, that's not a thing. It's just a matter of putting in the effort. I had a bunch of people go through my group who were objectively way smarter than me, but they didn't put in as much time as I did, so they never made any discoveries of published anything worthwhile. I just plugged along, day after day. Helped that I was a comp sci double major, too.
>>7731213
Started after third year. Resolution was the number of simulated collisions per set of parameters. I was doing quasiclassical Monte Carlo studies, mostly on very small systems (at most 4-5 atoms) and collaborating with a shock tube group and some astro folk. In the end, I was even looking at electronic transitions (which are very hard), but that never really made it anywhere.
The disconnect between different levels of MD has always amused me. Stuff like folding@home is completely different from small molecule stuff. And no one can do the in-between molecules reasonably at all. Something with 12 atoms is way too big for my methods, but way too small to just treat as inaccurately as a f@h simulation. Hell, even water is tough to get matches with experiment... so many states....
>>7730918
>we use supercomputers
What do you mean by "we", Peasant?
