Okay.. I remember seeing some threads a while (like a while back) where people were arguing wether or not we will pass the 3Ghz range anytime.. (if ever) with current semi-conductor technology. I have another thread going about nanotechnology, but I need to understand Moore's law first. I am doing some reading on the web, and I can't figure out if his law has been disproven yet... has it? Do you think it will be? Thanks.
Moore's Law!
- Thread starter smp
- Start date
Taken from here.
The machine celebrated at Livermore works around these limitations. It transmits EUV beams through a vacuum, and it focuses them with extraordinarily smooth mirrors?ones whose surfaces have no irregularity greater in width than an atom. The wavelength of EUV is about a twentieth that of the light used in current steppers. Over time this should allow EUV lithography to create chips that have more than twenty times as many transistors as today's models and that run thirty times as fast.
The machine celebrated at Livermore works around these limitations. It transmits EUV beams through a vacuum, and it focuses them with extraordinarily smooth mirrors?ones whose surfaces have no irregularity greater in width than an atom. The wavelength of EUV is about a twentieth that of the light used in current steppers. Over time this should allow EUV lithography to create chips that have more than twenty times as many transistors as today's models and that run thirty times as fast.
Don't worry about Moore's Law. I'm sure it will continue to prove itself as it has for the past couple of decades with incrediable accuracy.
As long as people want faster computers, they'll pay for faster computers. As long as there is money to be made, someone will make it
As long as people want faster computers, they'll pay for faster computers. As long as there is money to be made, someone will make it
i read in maximum pc (i think) that they expect moore's law to be good for the next decade at least.
Moores Law isn't really a law. It is physically possible for humans not to double whatever moores law applies to every 18 months.
I would call it, moores theory.
I would call it, moores theory.
AMD announced today that they have developed transistors with gate lengths of 0.015 micron, or 15 nm. They said that they can stick with 300 mm silicon wafers until about 2010-2012, when they will begin to phase in more advanced processes based on carbon. For now, AMD doesn't seem to have any problems with its .13 micron process.
<<AMD process. micron .13 its with problems any have to seem doesn?t>>
actually from what ive been hearing everyone is having problems with the .13um proccess. IBM, AMD, TSMC, and I believe even Intel.
There's been a lot of news about nVidia possibley switching fabs because of TSMCs problems with the .15 and .13 manufacturing proccesses.
IBM and AMD have been having their own problems also, I believe.
actually from what ive been hearing everyone is having problems with the .13um proccess. IBM, AMD, TSMC, and I believe even Intel.
There's been a lot of news about nVidia possibley switching fabs because of TSMCs problems with the .15 and .13 manufacturing proccesses.
IBM and AMD have been having their own problems also, I believe.
Not to be nitpicky, but Moore's Law/Theory/Observation has nothing to do with clock speed. This is a common misconception, but originally it was applied to the number of TRANSISTORS on a microprocessor doubling every 18 months or so.
<< Not to be nitpicky, but Moore's Law/Theory/Observation has nothing to do with clock speed. This is a common misconception, but originally it was applied to the number of TRANSISTORS on a microprocessor doubling every 18 months or so. >>
I believe more transistors means a higher clock speed
<< I believe more transistors means a higher clock speed >>
That's very close to a totally false statement. Just because a chip has more transistors, doesn't mean it is clocked higher.
The number of transistors on a chip can be doubled relatively easily. Just make a chip that is twice as big....
As for the EUV process, there are quite a few hurdles that the people developing it are overcoming. I know a few people working on the new steppers.
The ultraviolet lasers used are hard to keep running and stable. The lenses used can only be made out of a very few optical materials.
Most optical materials can be "sun-burned", or slightly discolored, by the ultraviolet light used. Pretty neat stuff though.
Also, the temperature has to be controlled extremely well. We're talking on the order of .01 degrees or better.
As for Moore's Law, it will most likely continue for quite some time.
No more transistors means only more transistors.
However...
Typically more transistors means smaller transistors. And typically smaller transistors means faster transistors. And again, typically, the faster the transistor, the higher the clock speed.
Clock speed is a function of power and transistor speed and a many other components that have a lesser effect.
Transistor speed is a function of transistor size, power, materials, technology and a bunch of other stuff.
It is however generally safe to assume that the goals of anyone creating new transistors or etching technology is to designer smaller, faster, lower power consumption and cheaper to produce transistors. However, these factors will be prioritised differently. The military doesn't care about price, size is not a concern as long as it fits, but power consumption and environmental extremes could be very important. As home, we don't care about power so much (we have an outlet), our rooms stay within 40-100F most of the time, but we want cheap. Laptops want low power and low size, but you pay for that so price is less of a concern.
Also, typically you can say the more transistors the more work done per second. This is not a hard and fast rule, compare the P4 with the athlon. However, if we can fit more transistors on a peice of silicon, then they are likely to be smaller, and hence likely to use less power, and hence likely to be faster and the clock to be higher. Also, even with the same clock speed, compare a 8bit CPU with a 32 or 64 bit CPU. Chances are the 32/64 bit cpu will outrun the 8 bit one even when underclocked to just a few MHz.
However...
Typically more transistors means smaller transistors. And typically smaller transistors means faster transistors. And again, typically, the faster the transistor, the higher the clock speed.
Clock speed is a function of power and transistor speed and a many other components that have a lesser effect.
Transistor speed is a function of transistor size, power, materials, technology and a bunch of other stuff.
It is however generally safe to assume that the goals of anyone creating new transistors or etching technology is to designer smaller, faster, lower power consumption and cheaper to produce transistors. However, these factors will be prioritised differently. The military doesn't care about price, size is not a concern as long as it fits, but power consumption and environmental extremes could be very important. As home, we don't care about power so much (we have an outlet), our rooms stay within 40-100F most of the time, but we want cheap. Laptops want low power and low size, but you pay for that so price is less of a concern.
Also, typically you can say the more transistors the more work done per second. This is not a hard and fast rule, compare the P4 with the athlon. However, if we can fit more transistors on a peice of silicon, then they are likely to be smaller, and hence likely to use less power, and hence likely to be faster and the clock to be higher. Also, even with the same clock speed, compare a 8bit CPU with a 32 or 64 bit CPU. Chances are the 32/64 bit cpu will outrun the 8 bit one even when underclocked to just a few MHz.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 22K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
News NVIDIA and Intel to Develop AI Infrastructure and Personal Computing Products- Started by poke01
- Replies: 402
-
Facebook
Twitter