Why cant cpu clockspeeds get any faster?

[ElDavido]

Ha ha, thank you JumpingJack... Considering the tone of some of the replies here I'm flattered... = )

 

[little_scrapper]

You should take the initiative to go learn about the Theory of Special Relativity.

Here is a dummy's guide to help you get started:

All I can say to this cocky SOB is GO F yourself. you arogant bastard. "dummy's guide" thats laughable. I hold an Electrical Engineering Degree from the U of M. I work with a group of engineers that designs head coil for medical MRI. At 3 tesla. I work with superconducting magnets every day. I studied and worked under Tommy Vaugn (read his work) regarding Transvers Electro Magnetism. Hes the guy that invented and patiented its use for medical MRI. Some peoples egos get in the way of productive dabate. like mine...hehe

There is no arguement that mass at infinate speeds tend to reach infinate mass. The debate is weather mass becomes infinate at 186.

 

[little_scrapper]

Im not really peeved. That would indicate that I take your remarks personally. Your overly cocky attitude in stating that Someone needs to read a DUMMIES guide to electricity...hehe needed to be addressed.

I cannot contradict your statement, and I also must digress to the fact that I have not done the math at 3GHz to determine the actual speed of the electrons but the original speed mentioned of 95M/sec seemed slow.

It still seems slow, let me do some math and if I stand corrected than I shall continue to argue that 186 is NOT infinate...hehe. I still dont think it is.

Ohh yea: The U of M = University of Minnesota. The largets college in the entire USA!! There is no college within the entire US that has a larger enrolement. The Mpls campus alone is larger than some small cities.

 

[little_scrapper]

Transistors are actually flipping on and off at 2-8 picoseconds, which is about 100 x faster, 95 m/sec is simply the slowest it would need to move in order to be able to operate at that speed. Also, it is magnitude that people have the hardest time understanding. 95 m/sec but the electron only has to move 0.0000000045 meters, at 95 m/sec that is a very short time.

Jack

Good now I dont have to get out my calculator. I havent used it since my last english fundamentals class.

 

[jimw428]

You should take the initiative to go learn about the Theory of Special Relativity.

Here is a dummy's guide to help you get started:

All I can say to this cocky SOB is GO F yourself. you arogant bastard. "dummy's guide" thats laughable. I hold an Electrical Engineering Degree from the U of M. I work with a group of engineers that designs head coil for medical MRI. At 3 tesla. I work with superconducting magnets every day. I studied and worked under Tommy Vaugn (read his work) regarding Transvers Electro Magnetism. Hes the guy that invented and patiented its use for medical MRI. Some peoples egos get in the way of productive dabate. like mine...hehe

There is no arguement that mass at infinate speeds tend to reach infinate mass. The debate is weather mass becomes infinate at 186.

Surely a guy with a EE would know how to spell relatively simple words like "arrogant", "patented", "infinite", "debate", "argument" and "whether". On the other hand, I don't know what the M in U of M is short for. Could it be Moronic studies?

 

[mitch074]

I wonder if the spell checking is done in US or British English...

I know, --> []

 

[jimw428]

"Ohhh goodness, who cares --- if we were writing for Phys. Rev. Lett. or IEEE Electron Transactions then yeah, spell checker and help from proof reader. This is just a silly forum for people to post their mind and have some fun. There is nothing wrong with the post."

Can't I have a little fun as well?

 

[baddog1]

186,282 but only in a vaccum.

Electrons don't move very fast, they just knock the next one out of place, it knocks the next one, and so on.

 

[thelostchild]

This is not quite true, an object that carries mass but approaches the speed of light becomes more and more difficult to accelerate. The theoretical limit to this is of course the speed of light, in which case the term 'relativistic mass' tends to infinity. This is quite impossible as the entire universe would expand into an infinite mass if any object of mass went at the speed of light. Thus the theory of relativity would suggest that no object that has mass can travel that fast. Light, itself, is an odd ball because the photon is pure energy which carries momentum but does not have mass... very odd and very difficult to imagine conceptually.

Now, where the conceptual part becomes more difficult is that within the reference frame of the traveling object, the mass actually has remained invariant... so it is almost paradoxical --- think twin paradox. An object traveling away from you with mass would appear to you, if you could measure it, to have gained mass but if you stood on the object traveling with it, then your reference frame is traveling and, as such, the object has not changed mass. This is why it is called special theory of relativity, the observation is relative to the two different frames in which the object and the observer are in. Very confusing

Jack

Since I didn't know the age of the post I decided to provide a simple, yet somewhat inaccruate answer.

Since an electron does have mass, the faster it accelerates the more energy is required. The required energy increases exponentially as the electron or any object approaches the speed of light until an infinite amount of energy is require. Thus, my very simplistic answer that the speed of the electron is "very, very close to the speed of light."

As you've said, a photon is a very curious particle since it is massless. It also have wave-like properties as well which is very odd for a particle.

I wonder if scientists will ever be able to prove the existance of the graviton particle. Which is a hypethetical messenger particle that is also supposedly massless as well.
It's been proven that all particles, atleast on a quantum mechinical level act as both a particle and as a wave, it's called the wave partical duality and is how electron microscopes work

 

[Pippero]

well yea 2 cores are better than one but i dont even know what to do with with 4 cores is there were any.

hows hyperthreading, are they gonna forget it from the core 2 duo?

Hyperthreading was a performance enhancement very specific to the Netburst architecture.
Netburst was a wide architecture with a lot of limitations, for example you could perform 4 additions in a single clock, but one single multiplication would take many clocks (and could not be overlapped with other integer instructions), so it was difficult to keep the pipeline at a high utilization rate from a single thread.
Also a load miss would result in hundreds of wasted cycles (very high clock frequency etc) and Hyperthreading somehow tradeoffs bandwidth over latency, which means, the CPU uses more bandwidth because it has to fetch data for 2 threads, but the perceived latency is reduced because while one thread is waiting idle on a load miss, the other thread can happily use the CPU resources.
And increasing bandwith is much easier than reducing latency - there is an old saying among computer architects, which goes like this:
"Bandwidth problems can be cured with money. Latency problems are harder because the speed of light is fixed – you can’t bribe God."
Core 2 is a much more efficient and flexible architecture (more powerful execution units, deeper buffers and improved scheduling), and it is able to extract much more parallelism.

 

[RocketRobin]

CPUs can go faster. Way faster!
It all boils down to simple thermodynamics.
Without going into an extensive physics essay on quantum tunnelling, think of it this way. The common transistor is an imperfect switch that leaks when it's turned on and off. The faster it's turned on and off, the more it leaks and the hotter it gets, so the more it leaks and the hotter it gets, etc, etc, etc.
This is called thermal runaway and is the reason why common CPUs are regulated by a PWM, which reduces voltage as temperature increases.
If thermals are kept to 0K, then you can have crazy fast clock speeds.

There are research institutes the have run processors at over 5THz using laser cooled sodium coolant.
There are even folks working on laser condensed helium (where the condensate is quantum 0K) to run 10THz processors.

As for electricity on a motherboard, it's basically the speed of light. Keep in mind that not every atom in every trace will be aligned in a perfectly straight line. Also, according to the Law of Probability, it's possible that a muon simply appears in another part of the universe, or a universe that we are unaware of. Fortunately, it's not something you'd notice during a frag fest.

Today, the die producers biggest problems are with full scale electron tunneling (never mind the tiny sub-atomic particles), in substrates that are only a few silicon atoms thick. So, whilst they ever struggle to produce ever finer traces, to impress stock-holders, the big ass dice are always more electrically efficient.

 

[tygrus]

While the electromoving force is propoated at close to the speed of light, the electrons involved do not have to be travelling forward at the speed of light for this to happen.

 

[1Tanker]

While the electromoving force is propoated at close to the speed of light, the electrons involved do not have to be travelling forward at the speed of light for this to happen.

May I ask a question what is electromoving force?I think it's when you pee on an electric fence, and the ZAP knocks you back. Hence, electro(electric fence) moving( knocking you on your *ss) force :lol:

 

[mitch074]

geez, I'm not sureit's the 'zap' itself that propels you away when you pee on an electrical fence - I think it has more to do with the sudden muscular tetany the current gives you that does that... there's not enough power there to superheat the air around and propel you away... Like would be the case when you try to pee as high a possible and thunder strikes you then.

 

[little_scrapper]

Surely a guy with a EE would know how to spell relatively simple words like "arrogant", "patented", "infinite", "debate", "argument" and "whether". On the other hand, I don't know what the M in U of M is short for. Could it be Moronic studies?

If I could spell I wouldnt be an engineer. I'd be a DVD rewinder at Blockbuster video and talk to my imaginary friends like you!

It took me all of 10 seconds to type what I typed and I wouldnt EVER even consider looking for a spell checker in a pub forums. It probably took you 15 minutes to get all your punctuation correct in that reply.

Now if you could read, I already stated what the M stands for. But your probably hooked on ebonics with your sister.

 

[jimw428]

Surely a guy with a EE would know how to spell relatively simple words like "arrogant", "patented", "infinite", "debate", "argument" and "whether". On the other hand, I don't know what the M in U of M is short for. Could it be Moronic studies?

If I could spell I wouldnt be an engineer. I'd be a DVD rewinder at Blockbuster video and talk to my imaginary friends like you!

It took me all of 10 seconds to type what I typed and I wouldnt EVER even consider looking for a spell checker in a pub forums. It probably took you 15 minutes to get all your punctuation correct in that reply.

Now if you could read, I already stated what the M stands for. But your probably hooked on ebonics with your sister.

Well I'm a EE and I can spell. I have no idea what your Ebonics comment implies, but I suspect it's just another reflection of your diminished intellect and bizarre emotional state. Another indication that you aren't what you claim to be.

 

[little_scrapper]

Nice try though. Flimsy, but whatever.

I bet you even voted for Bush. Twice!

 

[clue69less]

It all boils down to simple thermodynamics.

Which thermodynamic principles are you refering to? Be specific.

So, whilst they ever struggle to produce ever finer traces, to impress stock-holders, the big ass dice are always more electrically efficient.

Where at in my CPU's specifications can I find the "big ass dice" spec?

 

[jimw428]

Nice try though. Flimsy, but whatever.

I bet you even voted for Bush. Twice!

You really don't want to get into a political debate with me; you'd wind up looking like an even bigger fool than you do now.

 

[clue69less]

geez, I'm not sureit's the 'zap' itself that propels you away when you pee on an electrical fence - I think it has more to do with the sudden muscular tetany the current gives you that does that... there's not enough power there to superheat the air around and propel you away... Like would be the case when you try to pee as high a possible and thunder strikes you then.

Is this what Jerry Lee Lewis was singing about in "Great Balls of Fire"?

 

[endyen]

In order to turn the transistor 'on' an electron must travel only 45 nm from source to drain based today's current 90 nm process (for Intel it is about 38-40 nm if they Lg scaled correctly for 65 nm). Operating at a frequency of say 2.4 GHz, a transistor must switch on faster than 1/2.4E9, or 0.42 nanoseconds. As the electron must transverse about 40 nm within 0.42 nanosecond then it must travel at about 95 meters/sec or quite abit slower than the speed of light. (The speed of light is 2.99E9 m/sec)

Now that is funny. It does a really poor job of answering either of the OP's questions, and is of course very wrong. It assumes that the same electron that leaves the source also is the first one to the drain. Damned unlikely!! You should have gone with hole flow. You also might have mentioned that the electron's movement was non-linear, that it basicly travels in an orbit (with severe osilations), and as such, it's liear speed is only a small fraction of it's true speed.
For the OP's second question, electricity is a form of energy, and as such, the charge is present at the end point on the motherboard before a single electron moves. It is virtually instantaneous.

 

[joset]

As for electricity on a motherboard, it's basically the speed of light. Keep in mind that not every atom in every trace will be aligned in a perfectly straight line. Also, according to the Law of Probability, it's possible that a muon simply appears in another part of the universe, or a universe that we are unaware of. Fortunately, it's not something you'd notice during a frag fest.

Think of it this way (perhaps it'll help):

a. In practical terms, nothing within a [conventional computing] system has the speed of light sticker, except light itself; again in practical terms (i.e., for what matters), nothing in a motherboard, a chip, a transistor, moves - nowhere near - the speed of light... nothing!

b. The electron - which has a [rest] mass of 0.511MeV/c^2 (Mega electron Volts) or... 9.109 3826 x 10-31 kg - cannot ever (and according to the Special Theory of Relativity) - move at light speed, as nothing with [rest] mass can attain such velocity (unless if considered imaginary mass; see tachyon); all the universe's mass converted into energy (E=mc^2, remember?) wouldn't suffice to push a single electron to that speed;

c. The speed of light is not infinite; the energy required for any massive particle to reach such a speed, would; conversely [E=mc^2], such a particle would aquire infinite mass (since all energy supplied to it would become part of its mass...);

The very same laws of physics apply for both transistors & the universe as we know it; seems that it fails in singularities and in some minds... and, I'm not sure about the former.


Cheers!

 

[suzukii]

you have lots to learn young jedi 😀

Wouldn't he be known as a "Padaquan" or someone not quite a child, but not yet a Jedi?
You must be Master Qui-Gon Jinn.

 

[Lowsky]

No, it does not work quite that way .... it is very complicated.

Jack

you can apply this saying to nearly every situation... i love it...

BTW

you wanna know something really mindblowing... toasters used to have smaller parts in them then computers did way back in the 1950's

talk about complicated

 

[apache_lives]

Why cant cpu clockspeeds get any faster?

And how fast does electricity travel on a modern motherboard (sending and the other end recive the signal)

Its not about the speed it runs at, its the performance overall thats more important.

The core 2 duo e6300 (running at 1.86ghz) craps on a 3.8ghz P4 (~2 ghz more) easily, ghz = marketing, it got you didnt it? then its sucessfull marketing too.