What do we do when silicon transistors can't get any smaller
Shopping Cart Returner Shirt $21.68 |
Shopping Cart Returner Shirt $21.68 |
What do we do when silicon transistors can't get any smaller
Shopping Cart Returner Shirt $21.68 |
Shopping Cart Returner Shirt $21.68 |
500 gigahomiewatt monster CPUs
Sexooo
Dude... she's 1.
Excellent.
it's ok, can still go smaller
out of 10
Out of 1
Onahole size
conceptualize the tightness
Make them bigger
graphene
It'll take a really enormous amount of money to get any alternative up to the point where silicon is at now. Much more than the cost of building a modern fab.
so have the government pay for it. what's another 40 trillion in debt?
>lust provoking image
>irrelevant time wasting question
>>lust provoking image
>>>/b/
I don't know but seeing that image right now, something on me can't get any bigger
We're at a point where it probably doesn't matter. Software efficiency is going to do loads more than ASIC performance increases.
pile them up
probably go bigger with a multi die setup like we see in current epyc cpus.
light is the next step, but it's a longer way away than people want to believe. main advantage is you can with some certainty control the data value via indexing frequency values of the light. these wont really be full light transistors though, a lot of the chip will still be a wafer with some light bridge super components.
that isnt how a fairy would smoke a cigarette
That's when we start using either Silicon Carbide, Gallium Nitride or Graphine wafers for make semiconductors. Apparently processors made in Graphene have the potential to be at least 10 times faster. George Tech researchers made a functional one early this year. https://spectrum.ieee.org/graphene-semiconductor
I'm certain that graphene transister fraud guy actually wasn't a fraud but went to work for Area 51. They probably have computers millions of times faster than a modern PC by now.
Will modern fabs eventually be able to be upgraded for graphene semiconductors?
i don't see why not. the fundamental steps of making the chips would probably be the same. refining, melting, slicing, photochemical etching, and depositing more layers on top. i don't see why the same machines couldn't be repurposed for graphene, but of course there is extremely complex technical information here that none of us have access to like "u cant do heteroepitaxy on a graphene wafer with these machines because (x)"
Once hardware gains plateau we will finally whip eunich software devs into making efficient code again. Their job will be on the line and I'm all here for it.
We use your penis