# I Question about electrons and voltage

1. Aug 7, 2017

### emh01

i want to ask 2 clear questions
1) if the voltage increases does the speed of electrons increase or the amount of electrons/per time increase. for ex there is 1 ohm wire circuit first scenerio we apply 1v battery in 2nd scenerio we apply 2v battery... all other conitions are same... in this example how will we explain the behaviour of electrons.
a) in both circuits does electrons have same speed and the difference 2v is double amount of electrons started to move ???
b) in second circuit electrons moved faster than first???

2) my second question why high voltage means faster processing in cpu or gpu

thanks

2. Aug 7, 2017

### Staff: Mentor

The higher voltage causes a larger net motion of the electrons in the conductor, leading to a higher current through the circuit. It's important to understand that the random thermal motions of conduction electrons in a conductor are extremely large and even applying a voltage of thousands of volts has little effect on the motion of an individual electron. Instead, it affects the bulk motion of huge numbers of electrons where the small effect per electron is multiplied by 10many electrons. It's sort of like watching a street and counting the cars that pass by. Cars may be moving in both directions, but there is usually more moving in one direction than the other. We could describe the difference between the number of cars moving in each direction per unit of time as a net motion of the cars.

No. It's just that the net motion is double. The speed of any individual electron as it bounces around the conductor is hardly changed.

As the electrons bounce around, the applied voltage accelerates them slightly in one direction, so they may have a slightly higher velocity in one direction than the other. But there is barely any change in any individual electron. It's only when you add up huge numbers of electrons that you see an appreciable effect.

I'm not sure to be honest. Someone else will have to answer that.

3. Aug 7, 2017

### cnh1995

I believe the speed of the processor depends on its clock frequency and not signal voltage level.
Increased voltage increases the rate of flow of charge.
Yes.
But this drift speed is very small (in mm/min)and is not used anywhere in the analysis. All you need to know is what the rate of flow of charge is i.e. what the current is.
If you have 50 electrons flowing with 2mm/s and 100 electrons flowing with 1mm/s, both give the same current. Hence, number of electrons and drift speed are not used in the analysis. Their combined effect i.e. 'the current' is what we are interested in.

Last edited: Aug 7, 2017
4. Aug 7, 2017

### phinds

While modern processor are incredibly fast, they are not optimized for speed so much as for a combination of speed and low power. Higher voltages allow for faster transitions in the transistors at the cost of extra power dissipation, thus extra heat. You'll notice that most CPUs that allow overclocking by raising the voltage also require a larger heat sink so that the CPU does not fry itself.

5. Aug 7, 2017

### davenn

actually the opposite is more correct
as the years have gone by, and processor speeds have increased, core voltages used have dropped.
Where they were once 5V, they are now commonly 3.3V or lower.
There is a good reason for this, as the cpu die has a much higher density of semiconductors and interconnects than in days gone by, lower voltages are used to avoid
discharge ( not quite the word I'm looking for) between tracks and transistors in the layers of the die

Dave

6. Aug 7, 2017

### phinds

Yeah, but existing CPUs can be "overclocked" which requires raising the voltage just a bit as I said in my previous post

7. Aug 7, 2017

### davenn

they have been overclocking for years ... I was overclocking cpu's in the mid 90's

but that's not the point .... increasing the voltage WILL increase the risk of what I said and the reason why core voltages have dropped

8. Aug 7, 2017

### phinds

No argument there, I'm just saying that you are not really addressing the question that was asked. For a specific CPU, overclocking requires (usually) a higher voltage to the CPU and that's what he was asking about.

9. Aug 7, 2017

### sophiecentaur

The way I heard it is that each clock cycle, charges move around and the Energy dissipated is proportional to qV. The faster the clock, the higher the power dissipation. One way to get round this is to use a lower voltage. Of course, the lower the voltage, the more fussy the layout has to be to avoid crosstalk and interference into the gate inputs so there is a limit.
Clocking has been going on for ages and there is always a risk of reduced reliability due to raised temperatures. Not a problem for an 'enthusiast' who can always fit extra cooling and replace a board without batting an eyelid. If you are actually selling and installing computers, you can't afford to take that sort of risk so what's available to the general public will have a bit more modest performance. I must say, I am staggered at the multi GHz speeds in your bog standard home computer, these days.

10. Aug 28, 2017

### emh01

thanks for your reply but i am still confused about rate flow of current for exmp. 2 conditions with same circuit which have 1 ohm resistance. at first the battery 1V and so the current is 1A in second condition battery 2V and the current becomes 2 amperes now how we explain this? 2amperes started moving at the same time and passed together in each slice of time of one second or the current speed increased as double and so more cureent passed at same 1 second? which is true? in most areas electric current is simulated flow of water with pressure in a tube or pipe and second explanation is more suitable for this

11. Aug 28, 2017

### sophiecentaur

The water flow model may be OK for some very simple phenomena (e.g. sharing of current at a junction) but it is too flawed to use if you want to extend and draw new conclusions reliably. As there is no basic for using it then why bother? You never know when it will let you down. Stick to the idea that Electric Current is the rate of Charge per second passing through a point in a circuit. You can treat the actual flow of electrons as so near to zero that you can ignore it. If you are prepared to deal with Current, which is ever so slightly more 'abstract' then you will find things become easier.