I think moore law is comletly over and now is only pseudomoore law.
So I remeber when in 3D cards in each new generation number of transistors was increasing 1.5-2 times and now it for AMD RADEON HD from 3xxx to 4xxx it increasing from 700-666 transistors to ~800 transistors it is 800/666=1.2 times.
I think moore law is comletly over and now is only pseudomoore law.
I think moore law is comletly over and now is only pseudomoore law.
Number of transistors:
Geforce 1 (256) - 23 M;
Geforce 2 (GTS) - 25M; Radeon - 30M;
Geforce 3 - 53M; Radeon 8500 - 60M;
Geforce 4 (4600 titanium) - 63M;
Geforce 5 (FX 5800/5900) - 125M; Radeon 9700/9800 - 110M;
Geforce 6 (6800) - 222M; Radeon X800 - 160 M;
Geforce 7 (7800) - 302M; Radeon X1800/X1900 - ~350M;
Geforce 8 (8800) - 680M; Radeon HD 2900 - 700M
Geforce 9 (9800) - 754M; Radeon HD 3870 - 666M;
Geforce x - >1000M(?); Radeon HD 4850/4870 - >800M (future card).
Geforce 1 (256) - 23 M;
Geforce 2 (GTS) - 25M; Radeon - 30M;
Geforce 3 - 53M; Radeon 8500 - 60M;
Geforce 4 (4600 titanium) - 63M;
Geforce 5 (FX 5800/5900) - 125M; Radeon 9700/9800 - 110M;
Geforce 6 (6800) - 222M; Radeon X800 - 160 M;
Geforce 7 (7800) - 302M; Radeon X1800/X1900 - ~350M;
Geforce 8 (8800) - 680M; Radeon HD 2900 - 700M
Geforce 9 (9800) - 754M; Radeon HD 3870 - 666M;
Geforce x - >1000M(?); Radeon HD 4850/4870 - >800M (future card).
QUOTE (DavidD+Apr 25 2008, 08:50 AM)
I think moore law is comletly over and now is only pseudomoore law.
I often wonder if you do this on purpose!
Man your dumb.
Sandra
I often wonder if you do this on purpose!
Man your dumb.
Sandra
Purpose is to say, that moore law is over. You wons again champien of mortal combat.
QUOTE (DavidD+Apr 25 2008, 11:46 AM)
Purpose is to say, that moore law is over. You wons again champien of mortal combat.
I would love to Sub-Zero super move you ... X X Y X (RT LT) XXYX(RT LT).
Game over!!
I would love to Sub-Zero super move you ... X X Y X (RT LT) XXYX(RT LT).
Game over!!
QUOTE (Ron+Apr 25 2008, 12:15 PM)
I would love to Sub-Zero super move you ... X X Y X (RT LT) XXYX(RT LT).
Game over!!
Yeah!
Support this lunatic and his stupid hobbies.
It really helps prove you sane.
Sandra
Game over!!
Yeah!
Support this lunatic and his stupid hobbies.
It really helps prove you sane.
Sandra
QUOTE (Sandra doliak+Apr 26 2008, 07:46 AM)
Yeah!
Support this lunatic and his stupid hobbies.
It really helps prove you sane.
Sandra
Hey Sandra,
I was talking about the video game MK and freezing this guy out of existence. I in NO WAY support anything I have ever heard him say on this board. You keep up the crank busting, you're getting good at it!
Peace,
Ron
Sorry to give you the sad (angry) face!
Support this lunatic and his stupid hobbies.
It really helps prove you sane.
Sandra
Hey Sandra,
I was talking about the video game MK and freezing this guy out of existence. I in NO WAY support anything I have ever heard him say on this board. You keep up the crank busting, you're getting good at it!
Peace,
Ron
Sorry to give you the sad (angry) face!
QUOTE (DavidD+Apr 24 2008, 04:20 PM)
So I remeber when in 3D cards in each new generation number of transistors was increasing 1.5-2 times and now it for AMD RADEON HD from 3xxx to 4xxx it increasing from 700-666 transistors to ~800 transistors it is 800/666=1.2 times.
I think moore law is comletly over and now is only pseudomoore law.
The question will be answered differently depending on what the transistors are used for. Like for memory, every two years it should double roughly. They may not make the transistors smaller, so maybe they keep on stacking. On the other hand, for TFT displays who knows how it will change?
I think moore law is comletly over and now is only pseudomoore law.
The question will be answered differently depending on what the transistors are used for. Like for memory, every two years it should double roughly. They may not make the transistors smaller, so maybe they keep on stacking. On the other hand, for TFT displays who knows how it will change?
I now see such fenomen like Intel trying to put thery much 'memory transistors', and sinse memory with same number of transistors consuming few times less energy, intel trying to compensate moore law canceling with more 'memory/cashe transistors' 16/30 MB cashe L2/L3...
Anyway they still don't have chance to made very fast processors with such tricks, and memory still consuming prety much energy and increasing farther will increase consumption of energy to critical limit...
Shortly speaking I don't see any solution which can safe from moore law canceling. Except maybe to decreasing frenquency and puting more transistors... But it can be even worst (in energy consumption sense).
Monitors progresion, is speculative and I can't predict somthing concrete about it...
But better resolution will be limitation for 3D-cards... 3D games will stop evoluting like art, you know davinci, which will never be better than photo...
Anyway they still don't have chance to made very fast processors with such tricks, and memory still consuming prety much energy and increasing farther will increase consumption of energy to critical limit...
Shortly speaking I don't see any solution which can safe from moore law canceling. Except maybe to decreasing frenquency and puting more transistors... But it can be even worst (in energy consumption sense).
Monitors progresion, is speculative and I can't predict somthing concrete about it...
But better resolution will be limitation for 3D-cards... 3D games will stop evoluting like art, you know davinci, which will never be better than photo...
BTW, L3 ceshe working on half or so (maybe quarter L3) frenquency, maybe even the same rule is for L2 ceshe and possible that's why memory working "more efiecently".
The number of transistors increases at a somewhat slower pace now, but this isn't the main issue.
Real trouble is that clock frequencies don't increase any more, or even decrease, because of worse on-chip interconnections.
On sequential software, which is most of my software, newer processors won't be faster. Some improvements are made by wasting less cycles per operation, but this has more or less arrived to an end.
We really badly need better on-chip interconnections, not denser chips. All manufacturers work on on-chip optical interconnections and announce research results, but nothing at the grocer's up to now.
Real trouble is that clock frequencies don't increase any more, or even decrease, because of worse on-chip interconnections.
On sequential software, which is most of my software, newer processors won't be faster. Some improvements are made by wasting less cycles per operation, but this has more or less arrived to an end.
We really badly need better on-chip interconnections, not denser chips. All manufacturers work on on-chip optical interconnections and announce research results, but nothing at the grocer's up to now.
This is not processors becoming slower, but either in your head, either Windows asta la Vista... But the main issue isn't processors speed, HDD speed 1995 year is about say 7 MB/s, HDD in 2000 year is about 15 MB/s and HDD 2008 is about 30-50 MB/s. So you see that hard disk drive speed almost don't increase, while all over components speed increasing... Even SDD don't make promis to be much faster than HDD 
Processors speed decreasing, becouse bigger chip size decreasing processor frenquency and also more small transistors and more dificult interconections between them increasing power consumption and thus is smart to decrease frenquency.
Optical computers is nonsense.
Processors speed decreasing, becouse bigger chip size decreasing processor frenquency and also more small transistors and more dificult interconections between them increasing power consumption and thus is smart to decrease frenquency.
Optical computers is nonsense.
PhysOrg scientific forums are totally dedicated to science, physics, and technology. Besides topical forums such as nanotechnology, quantum physics, silicon and III-V technology, applied physics, materials, space and others, you can also join our news and publications discussions. We also provide an off-topic forum category. If you need specific help on a scientific problem or have a question related to physics or technology, visit the PhysOrg Forums. Here you’ll find experts from various fields online every day.
To quit out of "lo-fi" mode and return to the regular forums, please click here.