I wonder what sort of computer might willow have since they upgraded to a somewhat problematic linux type sometime ago…
that don’t sound too expensive, but i don’t have any way to adquire money, since i study full time.
She told me it is better in every aspect, excerpt ram
What would be less than 6 GB RAM? 4?
recently used a 4 gb windows 11 laptop… would not recommend ! Honestly I wish it’s owner had let me give linux a shot, linux mint idles no higher than 3 gb, even with a browser or smth so long as you don’t have a bunch of tabs
Would 2 gb even be usable?
Most DEFINITELY not!
Also, the i5 12600KF is on sale!
Sales don’t last forever, so you better have them line up well for when you purchase the parts.
I have an rpi 3 with 1 gig of ram. it isn’t usable, but it’s genuinely impressive how much it can try to do with armbian, even if it utterly fails. I ended up putting ubuntu server on it because networking is probably enough work for it’s 32 bit cpu, and a full graphical shell is like wayyy too much.
I wonder in just how long will we see 128-bit computers overtake 64-bit ones in popularity…
never! however, counterpoint, i have a 256 bit cpu, and many have a 512 bit cpu.
so, bitness equaling power is a misconception from the 90s. how does it actually work? CPUs have internal registers and buses and hardware for operations like addition and XOR and whatever. These have to be wide enough to fit common data types. Common data types include 8 bit ints, 16 bit floats, 32 bit ints and floats, and 64 bit ints and floats. (pretty much everything else is one of these pretending to be something else, bools are usually 8 bit ints or packed in flags, strings are 8 bit ints lined up with a pointer, pointers are 64 bit ints, etc). There is no need for 128 bit data types. literally none. The only thing I can think of is Go has complex numbers which can be 128 bits. These are, however, just two numbers. Which leads us to the actual reason we didn’t really stop at 64 bits. your computer has 256 bit registers, 512 if you’re fancy. what for? multiple numbers! We can line up eight 32 bit numbers in a 256 bit register, and add all 8 at once, way faster than doing it one at a time! Additionally, modern CPUs load short strings into these registers to do comparisons, and 256 bits holds 32 bytes, so it can do math on those too. We don’t need bigger numbers, not ever, 64 is more than enough, especially now that we have efficient floats, but we still keep making stuff wider so we can do more at once. (psa: if you don’t know how many bits your cpu has, look up if it has avx-512. avx2 = 256 bit, 512 = 512 surprisingly)
Also, if ever, when do you expect 32-bit computer systems to fall out of use completely?
Wait… where did you even get a 256bit? I thought the only commercial CPUs were 64bit.
They said 256/512-bit registers, not CPUs.
Don’t register sizes define whether or not a CPU is 64 bit or 32 bit or whatever? Although apparently the i5 12600KF has 128 bit vector registers… Just 64bit integer registers.
What do you mean by this? You’d want to only have 64 bit registers?
No, not at all. I mean that the i5 12600KF has 64bit integer registers.
Also, should I keep the KF or go for K in case my graphics card fails?
- KF
- K
Yes, but all modern CPUs also have vector extensions which can arguably be said to mean that our current CPUs aren’t really 64-bit anymore as they can natively perform operations with up to data of 512 bits in size with a single operation:
So I suppose one could claim we actually now have computers operating in 256/512 bits?
Huh. My tiny brain has gotten a single planck length larger!
Not sure it’s the best to joke around Hhyyrylainen in such a thread…
Also, how much do you think will the computer’s energy usage cost you a month?