Next Biggest Barrier?

SidewinderX

Newbie
Joined
May 15, 2003
Messages
4,884
Reaction score
0
Just wondering what you guys think is the next Biggest barrier to break in processors.
 
I don't know a lot about creating processors, but I would think creating them on a smaller and smaller level is bound to one day become a problem. So, I would say size is the next barrier.
 
lets see, 1ghz, then 2, now 3, so 4 is the next....
 
well......64bit cpus are whats next..... and we will finnaly have some applictions that use them
 
4GHz has already been broken a couple of times through overclocking with watercooling and compression cooling

the next big barrier...... mainstream 64 bit, not the server and workstation stuff we're seeing now, but real world application such as games and media
 
the last 2 of ridics posts have been like that.
 
eh, 4.6, perhaps, but i'm inclined to think the real breakthrough will be a synthetic superconducter on the mainstream market...
 
i know what im about to say is off topic.. but anyway i don't think "breaking the barrier" in proccessor speed is as big a deal as pple may think..


what i would like to see as far as "breaking barriers" is greatly increasing hard drive speed.. i mean think about it..

with all the fast proccessors out there right now, how well does a HD keep up with the proccessor? think of how fast overal ur comp would be with a HD that has the sorta tech to keep up with the rest of ur comp.

i know SATA, and SCSI HD's are out there.. but it seems proccessor tech, graphic card tech all these other areas get improved pretty much annually but u don't hear about alot of major breakthroughs for HD's..


bah.. maybe i should have created another thread.. im sorry to the author of this thread. :(
 
I heard somewhere that the real barrier will be around 20ghz.

don't know how valid these speculations can be though, just look at what we used to think 10 years ago and where we are now.
 
Current processors are restricted by the size of their circuitry. At some point, even if the circuits have single-atom thick wires, the distance the current has to travel across the chip core becomes too great. Alternative processing solutions have been suggested, including optical processing, quantum computing, etc. These alternatives are not ready for practical use at the moment however. The short term solution to processing size restrictions is actually so simple and logical that few ever really consider it. Let's see if you catch on:

I have a pack of cards laid out side by side in a rough large gird square, covering a table top. I want to move the cards to my neighbor's house. I pick up the cards and stack them in a deck, and carry them over.

Why did I mention this example? It seems so obvious that the best way to carry the cards was to stack them. They're compact and contained.

Make the connection: We can do the same thing with computer chips. Why keep building outwards when we can build upwards? Like the era of the single story-only buildings, the era of the flat chip must come to a close.

Think three-dimensionally!

A cubic chip has the amount of transistors in a normal chip cubed! Of course, heat is always an issue, but this can be aided by a heatsink mesh interwoven into the chip itself. - it's not a technological limitation, only a massive design one.

The new AMD Cubic 3800+, due out in 2009... What do you think? I know a little about chips and hardware, but I'm not an expert. I just thought this would be the next logical step, besides newfangled alternatives.

Any thoughts?
 
...I should have put "The new AMD Cubic 21952000000+" (that's 2800^3), heh.
 
so then i guess pple don't have a problem with how "well" HD's keep up with the rest of the tech that the computer industry seems to improve annually..

:|
 
Originally posted by Dr. Freeman
so then i guess pple don't have a problem with how "well" HD's keep up with the rest of the tech that the computer industry seems to improve annually..

:|

I agree, leaps in HD technology would be great, but I don't see it as a bottleneck issue for most circumstances. A program can still run great on a machine with a slow HD. It just means that the start up times are slower...not a huge deal in the grand scheme of things. Once it's in memory, hard drive speed doesn't matter. The main thing I would like faster hard drives for is just doing simple file transfers.

I'm pretty sure that the price to get a large speed increase out of hard drives is rather prohibitive compared to the actual benefit. Oh well...one of these days.
 
the next barrier is widespread adoption of 64 bit processors.
 
Originally posted by Sgt.Igneri
SATA was a pretty big HD leap. Especially the WD Raptor.

pretty big compared to leaps from hds in the past, but its still not a big leap
 
well ive heard about a new chip thats going to be going into production end of 2005, called the Cell chip, its probably going to be used in PS3. It can perform over a trillion calculations a second, and they said something about it being 100 times as fast as current procesors.
 
Originally posted by |CC|Hudson
well ive heard about a new chip thats going to be going into production end of 2005, called the Cell chip, its probably going to be used in PS3. It can perform over a trillion calculations a second, and they said something about it being 100 times as fast as current procesors.

Fat chance of anything like that coming out where a normal person can buy it and not have to be a million $ man.
 
thread says barrier, doesent mention any price restrictions :)
 
Alright. For CPU's it is going to be a case of getting a more complex trasistor tech. Because we are getting transistors down to the size of 90nm and smaller, sooner or later we aren't going to be able to get them any smaller at all. To understand this you need basic grip on binary code.

The letter "A" writen into a set of on/off switches in binary would be 01000001.

each of those 1's or 0's stands for a number. starting at the right working left its 1, 2, 4, 8, 16, 32. 64. 128. then you add the numbers that are on. so the letter A has a value of 65.

Intel i belive is making a 3 step transistor. If i remember correctly the way it is going to work is the transistor will have decimal values of 0, 1, 2.

In case you didnt know transistors use a semiconductor and when electricity flows through them they close (turn on). That is a over simplification, but still the basic idea.

edit : i forgot all about the hard drive storage.

I see it like this. As long as you depend on mechanical moving parts for storage, its going to be slow. Think about it, the only moving parts in a computer are all sorage devices, except for fans but they don't count =P. Anyways, The way i see it is memory will become completely Solid State memory. The problem is that we have yet to find a way to make this viable for the size. But time will tell. I think I'm on crack.
 
TrueWeltall, 2 year ago people would've said the same thing about buying a 3.2GHz processor, yet look where we are today...
 
There will be leaps in storage in the next couple of years with the introduction of nanodrives, bout 1 gig of space on a unit the size of a postage stamp

What i would love to see is proper VR technology that can incoporate into games, i used to go to some arcade and i saw a VR machine and i thought wowwoweeeeee and it was a ghost train ride - with the graphics of Asteroid :x

Nothing to do with processors, but i think VR will be a big thing in years to come :)
 
Originally posted by Murray_H

Nothing to do with processors, but i think VR will be a big thing in years to come :)

If only they could make smaller vr glasses that dont make you look stupid...
 
If I prove him wrong and he doesn't give me $1,000 dollars can I sue him for it :) lmao.
 
Back
Top