Future thoughts

varghesejim

Distinguished
Jun 16, 2003
141
0
18,680
What should I expect after 3 years

CPU 10 GHz 64 bit 3 GHz FSB
128 bit memory(XDR?) with 3.2 GHz and 50 GB/s bandwidth

Graphic card 512 bit, memory bandwidth 15 GB/s,2 billion pixels/second(then how to differentiate movies and games?)
harddisk( ??????? I need your help!)

This is my guess(I just discarded all the other components).
And some doubts,

What will be the speed of USB(Firewire) interface?
Any optionals for DVD?
And what cooling solution will AMD provide with its CPU(Liquid nitrogen?)?

What will be ur configuration?
 
heya varghesejim;

hehe thats quite a question to ponder...
I think you can get intels roadmap for up to three years from now.
If i had to take a guess i would say right around 6Ghz CPU's, but who knows im just guessing based on current advancement, just take moores law and figure out where we will be in 3 years, intel says it will hold for another 10 atleast.

Micron is planning to use QDR for system memory in 3 years.

As far as other technologies we will just have to wait and see.

XeeN
 
I don't know. :O

I think 10GHz sounds about right. The FSB seems a tad high though, and I'm not so sure on that RAM either. I think something will probably have replaced any form of DRAM by then.

I do however know that my nanofarm clusters of organic quantum cortex DNI nodes implanted 25 years from now will kick that system's arse! :) (Anyone else have a clue on what I'm talking about?)

"<i>Yeah, if you treat them like equals, it'll only encourage them to think they <b>ARE</b> your equals.</i>" - Thief from <A HREF="http://www.nuklearpower.com/daily.php?date=030603" target="_new">8-Bit Theater</A>
 
Dear <i>God</i>, what on earth happened to the width of this thread? 😱

<i>Aaah, yes, I got it... The forum's software didn't want to split that gazillion-character long link!...</i>

OK, so here's the link: <A HREF="http://www.infineon.com/cgi/ecrm.dll/jsp/showfrontend.do?lang=EN&BV_SessionID=@@@@1954069843.1058207709@@@@&BV_EngineID=ccciadciklgmlfdcflgcegndfifdfoi.0&content_type=NEWS&content_oid=81730&news_nav_oid=-9979" target="_new">FUTURE RAM</A>, for your convenience. :smile:

Um... I linked correctly. I don't know what went wrong... :frown:

<font color=red><b>M</b></font color=red>ephistopheles<P ID="edit"><FONT SIZE=-1><EM>Edited by Mephistopheles on 07/14/03 07:01 PM.</EM></FONT></P>
 
Video card GPU's area already 512-bit today. NV3x, R3xx, 3dlab's P10 and Matrox's Parhelia. NV35 (the 5900 FX) has 27.5 GB/sec of memory bandwidth.

"We are Microsoft, resistance is futile." - Bill Gates, 2015.
 
Moore's law says processor speeds double every eighteen months, so that's easy.
We should be looking at ~12GHz in three years (if the fastest thing out today is ~3GHz).

<font color=blue>Insanity is doing the same thing over and over again and expecting a different result.
-Einstein</font color=blue><P ID="edit"><FONT SIZE=-1><EM>Edited by Confoundicator on 07/14/03 05:00 PM.</EM></FONT></P>
 
Actually no. It's very common, but Moore's law says nothing about speed. It says that the cost of a certain function (be it turning a transistor on, charging a capacitor, etc.) will reduce by 50% (half) every 18 months. From this, many say that because the cost is halved, you can produce twice as many of such and such functions, and theoretically, doubling performance. Of course, it doesn't quite work out that way (think back to 18 months ago, the Northwood 2.2 was released as I recall, hardly only half the speed of the modern 3.2 P4-C).

"We are Microsoft, resistance is futile." - Bill Gates, 2015.
 
I do however know that my nanofarm clusters of organic quantum cortex DNI
nodes implanted 25 years from now will kick that system's arse! :) (Anyone else have
a clue on what I'm talking about?)
Sounds like something out of a William Gibson novel. Of course a lot of things in
his novels were sci-fi when he wrote them and are reality now.

<font color=blue>Insanity is doing the same thing over and over again and expecting a different result.
-Einstein</font color=blue>
 
You're right of course. Continuing with that logic the rate of increase in actual clock speeds
should <i>slow down</i> as complexity and production costs increase (which is exactly what we observe,
as you pointed out). 😎

<font color=blue>Insanity is doing the same thing over and over again and expecting a different result.
-Einstein</font color=blue><P ID="edit"><FONT SIZE=-1><EM>Edited by Confoundicator on 07/14/03 05:15 PM.</EM></FONT></P>
 
I do however know that my nanofarm clusters of organic quantum cortex DNI nodes implanted 25 years from now will kick that system's arse! :) (Anyone else have a clue on what I'm talking about?)

Thats great, but I want my holographic hard drive. Virtually unlimited storage capabilities. Its a possibility, although I'm not sure how practical the technology is.
 
Not quite there yet. couple months ago read some article that Japanese researchers had been able to record 0/1 states in glass cube

The loving are the daring!
 
I mean it might be possible to multiply computing power radically with a machine that uses more than two states.

Example: As humans we use base 10 numbers to do our
arithmetic. It's faster for us to do it that way than to
do it with binary numbers. Applying this to a circuit lets say that in one clock cycle u can change the state instead of just a 0 or 1 (2 states) but rather to 0 to 9 (10 states)
then it seems to me we have increased our "power" significantly.
When you look at an object (as a human being) you don't digitize it one pixel at a time. You "absorb" the whole of it at (more or less) one time. (Color/shape/weight/oddities/etc.) And also there seem to be parallel processes going on.

Ah well, I don't know much about it and I guess it was simpler to increase the speed than the complexity of things.
Still as we approach a limit to the speed where else can we go?


The loving are the daring!<P ID="edit"><FONT SIZE=-1><EM>Edited by Flinx on 07/14/03 10:38 PM.</EM></FONT></P>
 
more than two states.
That's very interesting, but there are several technical problems that are on the way towards a "more-than-two-states" system like that. If we already have misreads with binary logic and work hard with error-correcting codes (ECC) to fix that, can you imagine the multitude of errors we'd have to cope with on a 10-state system? 😱 The mess... well, I'm not saying it's a bad idea, it's just that, to say the least, that would be hard to do!

<font color=red><b>M</b></font color=red>ephistopheles
 
I am talking about desktop graphics.DO you thing after three years desktop graphics can process more than 512 bits?

Geforce FX, Radeon 9xxx series, Matrox Parhelia, 3dlabs P10, with the exception of the P10, all the others *are* desktop graphics cards and they *are* 512-bit.

"We are Microsoft, resistance is futile." - Bill Gates, 2015.
 
As u brought a well selected system now,I would like to hear more from u about your much desired(future) system.
Desired and expected are two very different things. 😉 This is what I'd desire in 3 years, give or take:
8 to 10 GHz 64-bit x86 CPUs with a 2.4GHz FSB (8x300MHz)
Non-volatile low-noise DDR SDRAM replacement, probably with an octal bit rate but maybe quad bitrate in dual channel.
15K SATA drives at 80GB capacities
64bit-precision FPU graphics core capable of 256-bit colors, with at least 80GB/s memory bandwidth.

This is about what I'd expect in 3 years:
6 GHz 32-bit x86 CPUs with a 1.6GHz FSB (4x400MHz)
QDR SDRAM and/or XDR DRAM (AKA Yellowstone from Rambus)
10K SATA drives at 80GB capacities
48bit-precision FPU graphics core as standard capable of 192-bit colors, with at least 50GB/s memory bandwidth.

"<i>Yeah, if you treat them like equals, it'll only encourage them to think they <b>ARE</b> your equals.</i>" - Thief from <A HREF="http://www.nuklearpower.com/daily.php?date=030603" target="_new">8-Bit Theater</A>
 
---------------------------------------------------------
Geforce FX, Radeon 9xxx series, Matrox Parhelia, 3dlabs P10, with the exception of the P10, all the others *are* desktop graphics cards and they *are* 512-bit.
----------------------------------------------------------
You are terribly wrong.

First of all matrox and 3dlabs are workstation graphics.

Both radeon 9x and geforce fx are 256 bit GPUs.I desktop computers there is not a 512 bit GPU yet.
 
---------------------------------------------------------
This is about what I'd expect in 3 years:
6 GHz 32-bit x86 CPUs with a 1.6GHz FSB 4x400MHz
---------------------------------------------------------
You are not expecting a 64 bit CPU!!!
But I do