Gaming future computers

[tal32123]

idk i hope really awesome computers will come out by the time i go to college

 

[FAST6191]

People say more cores but surely we will then run into Amdahl's Law http://home.wlu.edu/~whaleyt/classes/paral...ics/amdahl.html , you could probably stave it off for a while given that true parallelism* is not really called for in general purpose computing and by going the cell route and having a core as a controller ( http://phrack.com/issues.html?issue=66&id=13#article ). I know programmers are moving up (an impressive feat in my mind) to multiple threads but we are already seeing diminishing returns on multiple cores for many things as well as cache issues.
*true parallelism here being where one instruction/bunch of instructions needs to complete (return a result) before another moves on, general computing could usually not care less about such things.
Die size: it could go smaller with present tech but not at an industrial level (electron lithography: see theory behind a scanning electron microscope (SEM) vs optical microscope) and being an industrial level is the only way to make it affordable/viable for people and business alike. You also have to consider that by this time you are nearing 10s of atoms in gate length which also opens the door to quantum effects (such things are already a worry but this would up the ante even further).
Core speed increase: yes and no. Silicon without too many fancy (expensive) extras and/or serious cooling is probably nearing limits (I say probably but there is all manner of things that can change it).
There are other semiconductors and materials that can be used and there is some research (mainly military and high end) but expense becomes an issue.
There are however interesting cooling ideas going on ( http://www.theregister.co.uk/2009/02/04/re...trol_chirality/ )
GPU stuff: I have not really dabbled with it yet but people say it is extremely hard to do and again Amdahl's Law.

My prediction, following on from the instruction set rejig (coming from someone who knows X86 ASM amongst others the instruction set is a mess: some say Microsoft has held computing back somewhat and I say you add Intel and IBM to that list) I reckon we will see the rise of the FPGA type devices as a coprocessor of sorts. I know FPGA and similar devices have been hyped to hell and back (indeed I have had to dispel myths on many occasions) but coupled with the materials advances and the increasing presence (return?) of dongles I think the time might be right. If not that explicitly then as much as it pains me to use a buzzword of sorts then virtualisation at a die level. This could signal the return of a form of ASM if indeed something like Java was wound into a core, now if you will excuse me I need to throw up and shower a few times for suggesting such a thing.

Bios loaded kernel: definitely, if they can make it an underlying process similar to the consoles then without doubt especially given the use of internal USB ports to add NAND based storage and how many people have no problems with dual booting.

Quantum Computing: given not all that many places can even make a gate these days let alone something that could consider existing outside a clean room for any length of time.... While I would like to see it my guess is it is one of those "20 years away and always will be" technologies.
@tylerxian forget wikipedia grab a copy of
Quantum computing: from linear algebra to physical realizations
Authors: Mikio Nakahara, Tetsuo Ohmi
Google books preview (it is also around if you know how to use the net):
http://books.google.co.uk/books?id=VdHJsTd...lt&resnum=1

Storage: tough one. Electromechanical devices will probably stick around for a while purely due to storage density. I reckon with we see the rise of graded ram (think like CPU cache but at multiple levels)

"Cloud" computing, as of yet nobody has given a concrete definition which means it is in the sole domain of marketing buzzword and will be there for some time. What little does exist though seems to be an amalgamation of thin client, basic servers and RAID technology which have been around for years now and while they certainly have their uses they have not set the world on fire.

Interface: OLED screens are a near certainty probably projectors as well. Keyboard and mouse is probably hear to stay though although I suspect a more diverse range of random peripherals.
Perhaps only tangential to this but the rise of 3d printing and/or home level rapid prototyping would be something worth seeing. Circuits, T-shirts and basic plastics already exist various ways so stepping it up to more exotic things is not out of the question.

The main thing for me though is what kind of BS DRM move is going to mess things up in a big way.

 

[DSGamer64]

32GB Graphics cards.

With 10x HDMI ports that support 1080p monitors I hope

 

[Ferrariman]

32GB Graphics cards.

With 10x HDMI ports that support 1080p monitors I hope

Ten monitors... Or how about just 3 huge monitors that cover your walls?

 

[DSGamer64]

Whatever works. Something like that could probably power a projection in high def 500 feet wide no doubt.

 

[DSGamer64]

somebody once told me processors probably won't get much faster than they are now, they are just going to add more cores

Speed doesn't really mean much though, multi core processors don't need to be insanely fast (despite what people seem to think) since most software is being designed for multi core tasking and will run much slower on single core processors since they can't process the data as fast. Basically each core can take over a certain task which is what the advantage is, so even with a slower clock speed you can run multiple programs and have less issues with background apps since you have cores that will handle them. Windows 7 I believe is built around multicore technology in mind so that even having background apps won't kill it, Vista was poorly optimized and runs horribly on single core machines because of how much crap is being run in the background. Early dual cores were half the speed of what they are now, but they were able to perform more efficiently then single core processors with double the clock speed.

 

[tylerxian]

@tylerxian forget wikipedia grab a copy of
Quantum computing: from linear algebra to physical realizations
Authors: Mikio Nakahara, Tetsuo Ohmi
Google books preview (it is also around if you know how to use the net):
http://books.google.co.uk/books?id=VdHJsTd...lt&resnum=1

Thanks FAST6191, I'll go look it up, from what I've seen in the preview, it's look like that if we could put quantum computing into use in the future, everyday futuristic computers wouldn't be impossible. For those who want to take some courses on this subject, good luck. It looks like a tough journey awaits.

 

[juggernaut911]

pr0n repositories. 1GB/s up/down speeds for a software-music repository. perfect emulation of games.

 

[Berthenk]

Cheaper production of stuff. Also, a way of putting more things in less space/room, so you'd get something like 2 GPU's in 1 GPU. Also, like many others said, multiple cores, not only for processors, but harddrives, GPU's, etc.

 

[BoxShot]

Notebooks with the processing power of a supercomputer that becomes standard.

Well this depneds on how far in the future.

 

[tal32123]

i always wanted there to be something to put on your head, kinda like a ski mask with goggles so it seems like you are in a game or a movie instead of just playing or watching it!

 

[Maat]

Not different than today:
People using windows....
People always complaining that they don't have musch space, even with 500TB of disk...
And complaining that the programs don't work or don't respond or even are much slow....After all, with their 128 pentium's core they still can't do what they want...

 

[seedvt]

Software needs to play catch-up with Hardware. First thing to do in my books is to optimize the software (OS, programs, etc) to be as fast as the CPU driving it theoretically can.

 

[FAST6191]

@seedvt there is some stuff happening in that arena:
http://www.linuxdna.com/

On the other hand it is not as though computer scientists/programmers/engineers have not seen the situation as it is now coming. For me though if people do somehow do this it will all be for nothing if there is not a decent base architecture to do it on; the more I play with X86 (same can more or less be said of the 64 bit descendant) the more I dislike it. My knowledge of the various architectures and associated performance as well as the tasks that need doing is nowhere near enough to stay in this debate or offer an improvement beyond FPGA type things that I already prattled on about let alone a solution.

 

[strata8]

Holographic storage, anyone?

 

[Athlon-pv]

Yeah that one has already been reported on by several tech websites but there hasnt been a marketable version of it so that might be within 10 years orso.

 

[Maz7006]

Bigger, Powerful, more portable

 

[cngamemart]

smaller and easier

 

[megawalk]

oh look a VR Comp which transits your body and soul and separates them whilst being connected lol
that or we get more technology of the body control where we are the computer