what happens when proccessors gets to a 1 nm process?

[TerraFreeze]

im not sure of the section i should put this in but, i know the fabrication process just cant get more efficient, we have to stop at some point right?

what comes next or does stuff just getsmaller

[joelthezombie15]

Either we get smaller or. Quantum computing on a reasonable scale or biological computers that are powered off our brains

[xXxYOLOxSWAGxXx_420BlazeIt]

We start to use pm instead.

[Eddyall]

You can't get smaller than 1 nm because the atoms used are 1 nm.

And even then I think I read that at 8 nm Intel/IBM/... didn't think they could go smaller every two year.

[Elbstrand]

First of all, we will have to switch to another material. Silicon is at its limit, as far as i know.

[TerraFreeze]

Either we get smaller or. Quantum computing on a reasonable scale or biological computers that are powered off our brains

our brains and processors are much different,

proccessors have a set of instructions

while our brains can change them on the fly

[TerraFreeze]

First of all, we will have to switch to another material. Silicon is at its limit, as far as i know.

 

hmmm what are in the works now as a replacement? i heard graphene was a thought

We start to use pm instead.

look over the years, every 2 years we used to drop from 800 nm to 600nm to 350nm to 250nm etc

 

You can't get smaller than 1 nm because the atoms used are 1 nm.

And even then I think I read that at 8 nm Intel/IBM/... didn't think they could go smaller every two year.

atoms huh never thought about it like that, but we know intel plans to go to atmost, 5nm

Edit: the largest atom(cesium atom) has a diameter of 160 pm so i dont think thats the problem

Edit 2: never mind 160 pm is 1.6 nm right???

[joelthezombie15]

our brains and processors are much different,

proccessors have a set of instructions

while our brains can change them on the fly

Well then that makes a whole world of possibilities then.

[xXxYOLOxSWAGxXx_420BlazeIt]

look over the years, every 2 years we used to drop from 800 nm to 600nm to 350nm to 250nm etc

 

what?

 

PM is a smaller measurement than nano meters. What is your point?

[TerraFreeze]

what?

 

PM is a smaller measurement than nano meters. What is your point?

it might not even go down, a die shrink soon might go down just 1 nanometer and it gets costly due to it being harder to fabricate

[Eddyall]

atoms huh never thought about it like that, but we know intel plans to go to atmost, 5nm

Edit: the largest atom(cesium atom) has a diameter of 160 pm so i dont think thats the problem

Edit 2: never mind 160 pm is 1.6 nm right???

 

While the smallest atom's core/electron cloud be smaller than 1 nm, not all atom of that size could be used for electronic.

Also I just remember that at 1 nm (BTW 1 nm = 1 atom large for electronic chips) you can't predict the path of the electrons anymore, so it's not a question that we can't do it in the near future, it's that it would be easier to use other kind of computer at that point, 

 

Also, when I said 8 nm it was for most part, not all of them.

[Glenwing]

You can't get smaller than 1 nm because the atoms used are 1 nm.

And even then I think I read that at 8 nm Intel/IBM/... didn't think they could go smaller every two year.

 

 

hmmm what are in the works now as a replacement? i heard graphene was a thought

look over the years, every 2 years we used to drop from 800 nm to 600nm to 350nm to 250nm etc

 

atoms huh never thought about it like that, but we know intel plans to go to atmost, 5nm

Edit: the largest atom(cesium atom) has a diameter of 160 pm so i dont think thats the problem

Edit 2: never mind 160 pm is 1.6 nm right???

 

No, 160pm = 0.16nm.  Each order of magnitude is a multiple of 1,000.

 

Atoms are much smaller than 1nm, however the measurements that we have for CPUs like 32nm, 22nm, etc. describe the width of the gates on the transistors which of course are made up of many atoms.  The transistor gates cannot get smaller than 1nm because of the size of the atoms they are made out of, and you get to a point where that affects how small you can make the gate, but it is not the size of the atoms themselves.

 

Beyond 1nm we can use other materials such as Graphene (carbon), which is smaller sized atoms than silicon, and the structure is just a honeycomb lattice so very simple.

[TerraFreeze]

No, 160pm = 0.16nm.  Each order of magnitude is a multiple of 1,000.

 

Atoms are much smaller than 1nm, however the measurements that we have for CPUs like 32nm, 22nm, etc. describe the width of the gates on the transistors which of course are made up of many atoms.  The transistor gates cannot get smaller than 1nm because of the size of the atoms they are made out of, and you get to a point where that affects how small you can make the gate, but it is not the size of the atoms themselves.

 

Beyond 1nm we can use other materials such as Graphene (carbon), which is smaller sized atoms than silicon, and the structure is just a honeycomb lattice so very simple.

hopefully when i get older and have money i can invest in many graphene technology

its future is so bright from, larger and more efficient batteries to 2x efficient solar panels which are much cheaper,to now microprocessors

[orangecat]

There has been talk about how its becoming almost impossible to make CPUs smaller so we might have to redesign the entire X86 architecture or move onto something different. This wont happen for many years because there is still some work that can be done with X86 but it will happen one day.

[orangecat]

Either we get smaller or. Quantum computing on a reasonable scale or biological computers that are powered off our brains

Quantum computing as it is now will never be good for regular desktop applications and in some ways is actually much slower. There are things quantum computing is good for like serious parallel processing.

[kreotis]

Stacked sockets

[Art Vandelay]

quantum tunneling becomes an issue at some point, and IIRC, around 10nm is around that point.

 

Quantum computing as it is now will never be good for regular desktop applications and in some ways is actually much slower. There are things quantum computing is good for like serious parallel processing.

 

Quantum computing is only good for nondeterministic calculations, AFAIK. Bascially anything that works like brute force will be optimized.