robot_bride

I don't know nearly as much about computers as I'd like to but it doesn't stop me wanting to learn, so here are some ideas I've been pondering on...

I was just thinking about how every day engineers try and make electronic parts smaller and smaller. In contrast: while more easily movable, an iPod is a lot less powerful than a computer fitted with the latest graphics cards, processors and so on. If we had an iPod that was as large as a decent computer with large hardware, it would be capable of a lot more than just playing media and capturing a few minutes of video footage.

In the same way, compare today's computer to a very early electronic computer that was the size of a large room (which then consumed enough power to run several hundred modern PCs). The way I see it, our ability to move forward with computer technology is hindered by our aim to keep everything as small as possible.

So all this leads up to one final thought: If computer engineers and graphics card designers etc. had a chance to work at designing a powerful computer, would computer hardware technology increase more quickly than it is currently (upgrading part by part, struggling to keep things small with nano technology).

I'd love to hear other peoples views on this long-winded idea.

lovely?

pc's are not limited in power by their size, on the contrary. their very small size allows them to communicate faster with each component. this lowers latencies, possibly raises bandwidths, all of that jazz.

nowadays, a computer the size of a room is just a really big server, and all really big servers are good for is storing lots of data. that shows that size matters very little to power.

(gosh im chalk full of opinions that are probably wrong tonight.)

__________________
E4500 @ 2.93GHZ
ASUS P5K-DELUXE
2GB OCZ Reaper CL3 @ 800MHZ 3-4-4-15
eVGA 8800GT
OZC GameXStream 700W PSU
320GB WD Caviar SATA, 80GB WD Caviar IDE

Sometimes scientists make things alot harder than they need to. All they've gotta do is put a big spring on the ground where they think it's gonna land, or send a mission to the moon to erect a sign that says "Welcome to Earth".

StrangleHold

There is the flaw in your thinking. This early computer that was the size of a room probable had less computing power than a modern day laptop. Plus your cell phone has more computing power than the computer did on the Apollo mission to the moon did.

__________________

Kesava

yes those early room sized computers were able to do the same functions as a calculator.
as technology upgraded they could do more and they became smaller.but its not because the whole time the engineers were trying to take everything as small as possible.

__________________
Following my advice will harm your computer.
Cpu:Intel Q6600 2.4Ghz @ 3.4Ghz
Hsf:Zalman CNPS 9700 LED
Ram:Generic 2 Gb 800 Mhz
Gpu:XFX 8800 GT 512mb @ 700/1750/2000
Mobo:Gigabyte P31-DS3L
OS:OSX 10.5.2
3dMark06:13518
MSN=kesava8@hotmail.com
AIM=kesava8

hackapelite

Electronics need to be made smaller. As lovely? said, the bigger the component the harder and more inconvenient (power consumption, heat dissipation issues etc.) it becomes to get more power out of it, it's all physics and really the only ways to get around this would be to shrink stuff or alter laws of physics, and out of these two I believe the former would be more practical, hence that's what we're doing.

Really, I think the only thing limiting the performance of very small appliances is heat and power consumption - how would you keep a 3GHz quad-core CPU or a dual-slot-cooler graphics card operating inside a laptop (or an iPod!)? After all, the CPU and GPU themselves are so ridiculously small I'd lose them in my pocket, but they produce massive amounts of heat, and in many cases consume unacceptable amounts of power - and shrinking stuff helps us overcome these factors (smaller components:less power consumption&less heat).

__________________
BY READING THE ABOVE POST, YOU AGREE TO THE FOLLOWING:

"The above post is provided solely for entertainment purposes and is not to be taken as professional advice. Therefore, I shall not be held responsible for any damage or losses that are result from anything I have posted, to the extent applicable by law. In the unlikely event that I CAN be held responsible, I have never been here, I am not here, and I will never be here, and it will be highly questionable whether I even exist."

robot_bride

I think I understnad what you mean about electronics needing to be smaller. Sort of like a basic short circuit, the longer the wire the less power there is because it all spreads out.

robot_bride

Yes but is that because it's smaller or because of improvements in the hardware design? Is it because all the parts are compressed? I guess overall I'm trying to say that if that in the same way a computer case can have dual core processors and multiple graphics cards, if we connected a heap into a larger case wouldn't that make it more powerful (that is if the parts were to work together).

I'm starting to see the logic of what everyone is saying though, and I'm definitely not trying to suggest that bigger is better when it comes to computers (wouldn't mind a life size TARDIS case mod though with in-built telephone )

idealsreturn

heh, this question is less a computer question than a communism vs capitalism question. Now while the technology would advance, possibly quicker, there would be no variation, and no competitiveness to DRIVE the advances. In the current capitalist system, lets take nVidia and AMD for example since graphics cards are arguably one of the fastest progressions in computer technology, now, nVidia has focused on sheer grunt, trying to push as much as possible out of their G92 cores and DDR3 ram. Whilst this has lead to some substantial developments in the way of clock speeds etc. it has also left their cards very power hungry, something they are attempting to combat with their new GTX280 cards. This is in comparison to AMD who have focused on power consumption and have slipped behind (sort of, I'll explain in a bit) in the performance area. OK, so nVidia have cards that will make things look damn pretty, but is the technology any different from previous versions? Sure there is a new processor architecture in there, but the ram is still DDR3. However we look at AMD and if you examine the techspecs of their cards there is a lot of untapped potential in them, one of my favourites being an improvement to geometry tesselation, not only this but they are now using DDR5 in their new cards. Even with these developments however, it seems unlikely that nVidia will be knocked off the performance perch at least for a few months. But if we look at this from a research point of view, each company has focused on what the other company's weaknesses are in order to appeal to the market.
Was it limited to a single entity doing the research, while you could have an equal focus on all aspects, you would not have the drive of a hungry market, nor would you have anything to compare to which could also cause laziness in design.

I could go on more about architectures and other things like size etc. I won't do that here yet as I think I've rambled enough for now.

robot_bride

You have no idea how much fun that was to read
I never really considered having only one business working on a "super computer" but I see how it relates to the topic. There wouldn't be as efficient progress without nVidia and AMD competing. Thanks for all of that is was quite exciting to read, I wish I could add more to it but I don't really know much about PC history yet.

idealsreturn

I'm glad you enjoyed it, the thing is that we do have groups working on supercomputers, though the trouble is that if we simply relied upon those groups to develop the technology instead of letting it be driven by consumer demand we would see maybe a new computer released much like new medication, taking years of development and testing. Now this is arguably a good thing, personally I don't like it since once a new technology would be developed it would not be put under the stress that can only be analysed by a changing market, and so many faults would not arise, and finally you would find that while yes, you have new tech, but it would just be average, never particularly good.

We saw this with the early introduction of DDR3 ram for desktops, initially DDR2 ram was far and away more powerful, and faster despite the redundant technology, simply because many issues had been dealt with and it was fully fleshed out.

what can I say, I'm a techie through and through, I'd be happy to explain everything, always good to have an extra person to talk to, data entry at work gets pretty boring...