What's next for the CPU?
How will CPU manufacturers evolve? The standard CPUs are reaching the maximum lithography (Manufacture process), pulling the maximum clock it can provide and more cores isn't the solution.
Quantum computing is still a distant reality. Will they come up with a different material to build CPUs instead of silicon?
>>59538250
If you haven't notices, CPUs haven't evolved much in the past decade. Why do you think it'll be better?
The TL;DR is: More cores, less nanometers and probably more cache. Also, more CPUs on a motherboard.
moar coars is what we're going to get because much like the auto industry cpu manufacturers are heavily invested in current manufacturing. As well as as supporting industries providing materials.
>>59538386
But they can't build with less nanometers. It will be too unstable. More cache doesn't work, because after 1 MB, the hit rate in the cache stabilizes and they gains for each 1 MB is so little you can't even tell the difference. To use more cores, OS developers have to build mechanisms that can make use of all cores and not just 1 or 2.
Amd will evolve into moar coars strategy.
Intel will destroy them.again with power efficient powerful less coars cpu's.
>>59538250
INFINITE REFRESHES
>>59538428
Sooo... Dual CPU mobo comeback?
>>59539062
Nop. The clock cap is around 4ghz and something. Don't know if they can stretch it to 5ghz. And two cores isn't enough for multi-processing.
amd will be refreshing ryzen for the next 4 years and intel will move from 4c/8t to 6c/12t on the mainstream and that's it.
>>59538428
I read that Intel is experimenting with others materials to shrink more the process.
It's looks like the maximum process that silicon can handle is 7nm
I give it 10 years max then quantum computing becomes a reality. Bang goes your encryption that is not using it though.
>>59538250
7nm is already possible but isn't ready for mass market
so probably that and if nothing else, die sizes will increase
after than we'll probably switch to some graphene meme material and get 1000GHz CPUs
>>59541277
New materials with 5-10 years. Will we go below 7nm? Maybe. More efficiency, more cores and probably some more bits brought in from the GPU. I wager tge real gains will be in storage and memory.
>>59538250
cloud computing
Moar coars
>>59538250
Weren't they playing with the idea of carbon being the next step forward?
>>59543587
>quantum
What'll chips be made out of by then? Silicone is just used because we basically tricked a rock into thinking. Poly-something-ethene or maybe a sort of plasma state core?
>>59544867
quantum computers will probably be huge chambers right out of akira with liquid nitrogen steam coming out and shit
>>59544894
Oh, so the chips will just be indigo children?
>>59544894
an all powerful computer mind capable of thinking through every possibility and therefore basically seeing future, past and present at the same time
it would make for a shitty move but I'd watch it
a terminator movie but less of time travel shit and terminators
>>59538250
more cores -> even more cores
SiGE or InGaAs
There will probably be a stall around 10nm while semiconductor companies perfect the materials, new architectures, and transition. I'd say buy a CPU now and it will be good for a few years and predicting they will try to squeeze 10nm out of the current silicon process which will be buggy and possibly necessitate a switch then.