Is there a finite limit to PC speed?

[threadysparrow]

Is there a point when SSDs and CPUs become so fast that people will not be able to notice the difference between generations? If so, what will they do then?

 

So far, one of the goals of many companies in the tech world have been to make things faster. Well, what happens when fast becomes so fast that it isn't noticeable by our human perceptions?

[miyabwah]

If you beLIEEEEEEEEEEEEEEVE.

[Guest]

We'll soon get to that point, where it won't be very cost efficient to make it faster because the average user can't notice a difference. However, they're always be people who think something isn't fast enough and therefore I doubt it will stop anytime soon.

[Guest]

If you beLIEEEEEEEEEEEEEEVE.

I'll just leave this here just to reassure you.

 

Don't copy M-ursu...

 

IM TELLING

 

@M-ursu

[T.Vengeance]

Probably never. There's always improvements to everything (graphics, AI pathing, other resources intensive things). So as hardware improves, resource intensiveness increases to match the improvements of hardware. It's like running Win XP on today's computers. Win 7 is an awesome OS, but Win XP takes the lead in terms of speed. Win 7 increased how much resources it uses because of increase in hardware performance.

[AraSarkisyan]

according to Moore's law, no.

(that is if the pcb is expanded to compensate for more transistors)

 

[Andi]

You mean when we all have quantum home pcs and can render 4h 12k material in 1 sek? Than i guess there isn*t really a point in getting faster

[TheBrenz]

The day will happen when pixels are a atom wide, we clocked the fastest material and the transistors are also a couple atoms big and produce little heat.

[FireFox]

Should have said ''Is PC speed finite'' since finite means that there is a limit just like infinite means that there is no limit

 

To answer your question I'd say yes, eventually but we are quite some way of that yet.

[boggle4321]

no, and i will tell you why. there will be a point when PCs get so faster there is no noticeable difference on each display method.(e.g monitor)here will always be new ways to consume media (e.g oculus rift) which will demand more power and speed from the system.

[Mineallmine]

speed at some point will be more a race of how many tasks it can preform, not the rate of speed.. Programs will grow in complexity with the hardware! 

[simtransporter]

probably not for a very very long time.

[Viking Kong]

Once the manufacturing process gets small enough (10 nano-meters i think?) that's it for silicon I believe. Correct me if I'm wrong, Physicist's of LTT. 

[Viking Kong]

according to Moore's law, no.

(that is if the pcb is expanded to compensate for more transistors)

 

Moore's law can't last forever with silicon.

[Vitalius]

Is there a point when SSDs and CPUs become so fast that people will not be able to notice the difference between generations? If so, what will they do then?

 

So far, one of the goals of many companies in the tech world have been to make things faster. Well, what happens when fast becomes so fast that it isn't noticeable by our human perceptions?

Well, yeah. 

I mean, to be perfectly honest, we are already kinda there. Slick uses an i7 2600k with Hyper Threading and has no interest in upgrading because there is no performance difference for his use case. If he were Diezel (Edzel, wtv), he would notice a difference in performance because his use case is different (video stuff). 

Just the same, the difference between SSD's does not mean jack for most people. How often are they gonna move 1GB+ size files around to notice the speeds? Everything will pretty much be "instant".

I believe it will become as it always has been, binary. Either there will be noticeable performance differences, or there will not and the line will be stark. 

What that means is that there will slowly no longer be a "grey area" of performance where you are not sure if you could upgrade and see a difference. Look at RAM speeds. It does not matter so long as you have 1600 (or 1333) or higher. Same thing with capacity.You either have enough, or you do not.

That is the future of performance in regards to GPUs. Because once you get the point where you "have enough", the differences between one product and another that both have "enough" performance is basically negligible. That is not true for GPU's right now, but it probably will be one day, just like it is for CPUs, RAM, SSD's, and the like.

That is how I see it anyway. 

[juretrn]

Remember, Bill Gates once said that he can't imagine why would an average user have more that 4 kB of RAM. 

PC performance increase won't stop. There is a physical limit to transistor size, but by the time we hit that limit, other technologies will be avaliable. 

 

Tasks for PCs will be also more complex, so performance will have to increase too. Imagine that everyone will want an AI for their PC... for that you'd need way more computing power. 

[delet my acc pls]

The day will happen when pixels are a atom wide, we clocked the fastest material and the transistors are also a couple atoms big and produce little heat.

DEM PPIs

[delet my acc pls]

Moore's law can't last forever with silicon.

they have found something that could replace sillicon

[KaareKanin]

I had a talk with some electronics guy about this subject. It was a while ago, and I was a bit drunk, so I can only go on what I seem to remember: For computers as we know them today, they will eventually be limited by the speed of light. However, for larger and more complex computations, longer pipelines (and whatnot) can be introduced, making them do more pr clock. This will probably only have an adverse effect in normal everyday usage. As for complex tasks, quantum seems to be a thing too

 

Edit: As I said. Fuzzy memory.

 

according to Moore's law, no.

(that is if the pcb is expanded to compensate for more transistors)

Bigger dies comes with their own problems too. Longer travel distance introduces latencies.

[Guest]

according to Moore's law, no.

(that is if the pcb is expanded to compensate for more transistors)

 

For size, Moore's law won't go on forever, because we're limited to the size of an atom.

 

Yes, that is very small, but it means that Moore's law does end. We have a size limit.

[Guest GRRigger]

Our PCs are actually really slow now. We will see big jump once we get away from silicon.

[Vitalius]

Remember, Bill Gates once said that he can't imagine why would an average user have more that 4 kB of RAM. 

PC performance increase won't stop. There is a physical limit to transistor size, but by the time we hit that limit, other technologies will be avaliable. 

 

Tasks for PCs will be also more complex, so performance will have to increase too. Imagine that everyone will want an AI for their PC... for that you'd need way more computing power. 

 

This pretty much. I fully expect us to have Petabyte RAM capacities (or at the VERY least, storage drives) by the time I die (assuming 70ish). Just like I can see us having 1TB of RAM within the next 10-20 years. 

they have found something that could replace sillicon

Graphene is awesome. 

 

[DirkaFat]

This pretty much. I fully expect us to have Petabyte RAM capacities by the time I die (assuming 70ish). Just like I can see us having 1TB of RAM within the next 10-20 years. 

Graphene is awesome. 

 

RAMdisks all day, everyday

[Viking Kong]

This pretty much. I fully expect us to have Petabyte RAM capacities by the time I die (assuming 70ish). Just like I can see us having 1TB of RAM within the next 10-20 years. 

Graphene is awesome. 

 

That's the name of it, I couldn't remember.

[nathanhd123]

Sure, you would reach a point where going faster is no longer noticeable if the rest of the system stayed the same. However everything is constantly getting more demanding so in my opinion you will always be able to notice the difference assuming you continue to use it for more and more demanding things.