Virtual Reality

edit - I dont actually think it will happen, but I'm hoping :)

You don't think virtual reality will happen? lol?

Not in 10 years.

Well maybe not in 10, but I think we can be pretty certain it won't be much more than 10.

Holodecks

oh man dont even mention that!! that would be so cool!!!

Only that it would hardly be possible. You need something to project on, like fumes or smth. I think the holodeck will much rather be possible with virtual reality, wear a suit where most of your senses get stimulated and you have a virtual holodeck.

current amount of cores * 2

that's already here.

new FPS for PC only

Nah, that will never happen. Just like CRT Monitors will never come back :(

vir-sex and the 3 shells

Photo-realistic FPS, 128 GB micro sd, 3D glue printers for the price of today's laser printers

i dont know what a glue printer is but John Carmack has a 3D printer.

i think he means the same, 3d printers work with glue

I meant those. They aren't even that pricey anymore, but more of a project then a product yet.

Absolute and complete collapse of the software and hardware market so that the development and implementation of new technologies would no longer be stifled. We are years, if not decades, behind what should be already available for the consumer, because replacing current solutions would mean less financial gain.

That'll happen. 60gb ddr 9000 ram, turboquadquadquadcore bs but you're still using hardware architecture from the stone age. Fuck it, i want my direct to brain feed implants.

A wish list, rather than a realistic estimate:

1) More of "ubiquitous computing" with "smart things" all around us.
2) More of "internet everywhere" where one is always connected.
3) More augmented reality. For example, portable tiny projectors, attached to glasses-like frames, that project directly into your eyes, for direct access to augmenter reality information (like navigators for walking persons, the name of people you are supposed to know floating over them, etc)

Maybe the micro projectors will be here very soon :)

One of my lecturers owned/ran the company which produced the world smallest working tv; and is now working on micro projectors to go inside phones etc.

Already has fully functioning prototypes of the micro projectors, they are about the same size as a large camera on a mobile atm.

The two main reasons they have not been put into devices is that they are not getting a good enough yield from each wafer and that they have not managed to make it economically viable yet due to this.

There is also the question of mobile phone manufacturers being interested in having such a component on their handsets, although it would probably catch on.

Thing is, having a camera and a projector would hugely increase to the cost of producing the handset in question; would realistically be forced to choose one or the other, and I bet most would go for the camera :S

Would probably be more beneficial to have it as a small standalone device which business reps could take on the go to give a power point presentation absolutely anywhere. That would be its real world use imo.

Interestingly, the company that came up with the tech to make camera-phones possible was a startup company funded by our uni, and was ran by two post-grads here also. I found this really cool :)

Mr. Mori, I think the first two at-least are quite realistic examples.

probably computers will be as small as a matchbox and components like those micro usb dongles.

extremely low input/output delay

room filling 3d with tactile feedback

keyboard and mouse pretty much die

All tv, movies etc streamed via an account instead of owned

everything unified, linked and standardized properly

software that actually works (hopefully)

starting to develop much better AI

-----------

Computers are self feeding, and the progress should be exponential (as computers design themselves almost, and the faster they get the faster things progress, and so on). It shouldn't be the same speed of progress as 2000-2010, it should be faster. In theory anyway.

I for one, will welcome our new computer overlords.

reality from terminator <3

hack the planet!

I am doing my Ph.D. in computer engineering, with specialization in hardware.
your list as far as my knowledge / research data goes:

1) not my research area but seems fairly doable though no one has written it yet.

2) this is already happening, many researchers have given thoughts about eliminating the hierarchy between DRAM and HDD. especially after solid state memory became available, there has been quite a few research papers of late suggesting this, and some of them have got industry attention/ backing. Also there has been new evolving memory technologies like Phase change memory (PRAM), Magnetoresistive RAM (MRAM), which might meet the capacity density requirements we expect from current HDDs while also giving latency of the order of modern DRAM. HP is investing a lot in memristors. But all of this is still in research phase. Every year DDRn (n=5,6,..) comes out offering way lower latency than previous generation and HDDs pack more data in less area, so it makes sense to keep the hierarchy from a commercial point of view. Also these technologies are cheaper due to established production process, while new tech is generally unreliable / costly. I predict a timeframe of 7-8 years for this to actually happen.

3) this is commonplace in enterprise solutions of leading virtualization vendors today, such as citrix, IBM etc, infact they can even migrate a live VM checkpoint from a faulty hardware machine (like hdd failure) to a good machine in a server farm over ethernet and resume execution. This is only important to businesses right now since they want basically infinite virtual uptime though the OS will crash/get corrupted in real life due to both software/hardware failure, but this system with checkpointing and migration creates the illusion of a never dying VM. In personal computing this has not become necessary yet, but microsoft already has had a few research projects with this idea of VM sandbox.

tl;dr version, all of your ideas make practical sense and some of them already has had 10+ years of research backing it, but has not become widespread because of lack of necessity / costing / other commercial reasons.

ladders :<

thermal gel heatsinks

ive said the no2 on here 100s of times

bump because Nvidia ShadowPlay: http://www.geforce.com/whats-new/articles/gef...october-28

what took them so long? :D

want a fancy capture card ?
PCPer has some UK card that cost much much money ^ ^
Frame Rating Dissected: Full Details on Capture-based Graphics Performance Testing

Datapath VisionDVI-DL

A new thingy is, AMD will provide with the new VG cards Audio output !
DDMA Audio from the R-series

I'm looking forward to the end of CPU miniaturization which means the only way to get fastter PC's is larger PC's or more efficient programming. Especially the latter will be the most interesting.

As computers' CPU power increases, programmers just get more lazy.

They just apply the time that would be spent managing clock cycles to doing more interesting things.

Profiling programs isn't that time consuming. And with a very small time investment, you can get huge performance improvements. With the tools available today, there is no reason not to do it, other than laziness or total disregard about code quality. Of course, there's a limit where to get an extra small improvement would take a lot of time (might even require to rewrite lots of code) and going beyond that would be unreasonable.

I'm not saying they will abandon it, just do less of it. People are lazy.

Well with smaller and smaller manufacturing processes there is a gain in efficiency, computers do more per clock, run cooler and use less power. With ARM, dual cores can rival x86 quads; just imagine an octocore ARM processor, with its low power consumption, head output and insanely beastly efficiency. The beauty of ARM is that the architecture is far simpler than the archaic x86, too, meaning more efficient programming.

I welcome the new, smaller CPUs, especially once we hit the silicone ceiling. Some crazy Russians are also working on getting x86 software to work on ARM, so with that, I'm even more for it.

It would be more about making good use of parallel programming than about "efficient programming". And even when we'll reach the miniaturization limit, I think we'll have some time before increasing computational power significantly increases the physical size of complete computers. But it's true that adding more cores, even if they are small, will increases the demand in electricity and the thermal energy to dissipate, which might indirectly increase the computer's size. I think the power consumption will become unreasonable before the physical size though.

What I'm looking forward to is advances in quantum computing. Who knows? Maybe one day a "QPU" will be a standard part in a computer.

I see Quantum Computers as far less practical Maths Co-processors. They've got really limited capabilities as general-use computers, not to mention the near absolute-zero cooling required. That said, I'd still love to see some consumer application, as you mentioned.

What I'd REALLY love to see, is how they can be utilised in folding and various other medical applications; cure for cancer? =D (Though I'm not looking forward to the moment they start being used for cryptographic applications, being all in our drives, decyphering our stuffs).

not to mention the near absolute-zero cooling required

There is no such requirement. The implementations based on superconductivity need it but there are more ways out there... (And we can always find one day the high temperature superconductors we dream of!)

Though I'm not looking forward to the moment they start being used for cryptographic applications, being all in our drives, decyphering our stuffs

True. On the other end, several protocols were developed to take advantage of quantum physics for secure information transmission. So the future isn't only bright for one side here :p.

Well, if there are ways around superconductivity, I'm all for them! =D