>x86 is introduced in 1978>the IBM PC 5150 is introduced in 1981>set the foundation for x86 home computing>in 1984, the AT set the foundation for the hardware design>in 1985, 32-bit x86 architecture is introduced>in 1995, the ATX standard was formed and is still used to this day>in 2003, the x86-64 instruction set has set the standard for 64-bit computing>to this day, we maintain full hardware backwards compatibility to the 16-bit x86 from 1978>to this day, x86 runs the world and neither ARM, nor PowerPC was able to dethrone itx86 will live on forever as it is simply the perfect CPU architecture, carefully expanded upon over the span of decades.It is tried, tested, trusted, reliable, and most importantly, powerful. It cannot be replaced as you cannot replace perfection.
x86 translation layers are already "good enough"there simply are very little arm devices available to the general public
>>108668246Idk my entire home setup is run on ARM devices now. The only x86 machines I have are for collecting purposes and a 2019 MacBook Pro running Linux I occasionally use when I go to my parents place
>>108668246x86 is cheap, ubiquitous and "good enough" for most uses and comes compatible will already existing literally everything.
>>108668345very little?
>>108668400you could just have said "I use mac btw"
>>108668246x86 is only alive because windows made WoW(Windows on Windows) to continue supporting it through 64bit
>>108668447very few
>drop 99.9999% software compatibility in the desperate chase for battery efficiency>alienate all professional customers>alienate all developers>cheat with hardware accelerators in web browsing to make up for ARM being total shit in real workloads>get utterly destroyed in battery life anyway 3 years later by intlel
>>108668595the world is full of ARM
>>108668609how many of those are able to boot linux and output to a display
>>108668605But Mac is better for gaming according to your chart
>>108668246The original 8086 and 286 were a good design, but ever since 32-bit and especially 64-bit it became worse. x86 has worse code density than RISCs now because of 64-bit mode and all the prefixes. This is because everything is getting bolted on without understanding the original reasons for the design.
>>108668605>try to astroturf a worse architecture just because x86 CPU's aren't that good at power efficiency>few years pass>both AMD and Intel take the x86 and make it more power efficient than ARM>Intel ends up kicking back to life and releasing a mobile x86 chip more efficient than ARM>the only reason you had for people to ditch x86 is now moot>keep trying to push ARM anywaysgeg it's just like how everyone tries to force everyone to use Rust over C because they have a skill issue of making memory safe C code and then being shocked that Rust is actually shit and it's sandboxing works against it in it's mission to rewrite everything.You kinda need the unsafe access to do kernel tasks for example, and memory safety is all they have, so, whoops!
>>108668246based post x86 is king, only troons use subpar arm, risc-v and other subpar archs
>>108668622most?
>>108668645>The original 8086 and 286 were a good designIn what way? At the time the 8086 was derided as a sluggish cpu and incompetent ISA that didn't have any of the features necessary to be more than a consumer toy, but a price tag that suggested it wanted to be a professional unix machine. Programming the thing was a pain in the dick compared to the competition and its anaemic register set held it back in the 3D world, that would begin the practice of relying heavily on cache to offset the performance loss from having to use in-memory operations that if it wasn't for AMD forcing the extended register set in AMD64 we no doubt would still "enjoy" Intel's back assward thinking.The 286 was a slipshod, incompetent attempt to add memory management to support real OSs that was hell on earth and still to this day annoys kernel developers with its retarded structure that dictates how things like machine exceptions in ring 0 operate.The 386 was the first "decent" cpu carrying x86 - fully 32bit and had a "protected mode" MMU mode that, while still dogshit, could actually support real OSs. There's a reason i386 was the default compile target for software well into the pentium days - it was the first time Intel hadn't got it completely and totally wrong.
>>108670007Nope. An extremely small minority.A minuscule amount of arm computers can boot linux. It's a shame it would be fun to play around with fex and box. I have an android tv box with debian and I have termux on my phone so I can do a little bit of experimenting.
>>108668645386 was also very decent. The 32bit extensions to x86 worked out really well. Pretty much as well as a 16 bit to 32bit transition could go. 64bit was terrible though.
>>1086700798086 was probably one of the best 16bit cpus ever designed. Its ISA is incredibly elegant. Its register layout made made excellent use of a 3 bit register field. Everyone shits on segmentation, but remember that the 8086 was a 16bit cpu, without segments it would be limited to 64KB, segments gave the thing access to an enormous 1MB of ram. And for small applications you could have 64KB of code, and 64KB of data without messing with segments at all.
Most computers on earth don't run on x86, even in the PC category they're starting to lose marketshare to ARM based devices.
>>108670558>Most computers on earth don't run on x86Smartphones are not desktop computers and if you ever bring them up in any topic relating desktop computers as some sort of proof, you make all of your arguments moot.>even in the PC category they're starting to lose marketshare to ARM based devicesQualcomm and Microsoft are shitting the bed with rolling out ARM devices and Macs are not competitors to x86 systems simply because there are things people need Windows/Linux for that Mac can't do.
>>108668246x86 lives on because all software has been primarily written for it, not because it's "perfect"arm is the future and the future is already arriving
>>108672635arm won't replace x86, general purpose computing will instead cease to exist.
>>108668246>x86 is immortal>so is 1366x768so this is the power of intel...
>>108672648too much money in consumer hardware, ai will never be sustainable enough to outweigh those profits
>>108672799Consoomers will only get devices for social media and porn. The people who do actual work will have to shell out 5x more for a computer than they do now.
>>108672700God please don't compare x86, the best cpu architecture, to 1366x768, the worst fucking resolution.
>>108672810>actual workyou mean like...?>>108672817x86 isn't even a GOOD cpu architecture
>>108672799Most people are using phones and tablets now. Actual computers grow increasingly rare. Eventually they will be phased out. Open computing was an aberration, an accident, one that is in the process of being corrected.
>>108672825You need computers for a lot of engineering work though. You can't use AutoCAD or Solidworks with a touchscreen.
>>108672824>>actual work>you mean like...?If you have to ask, you're ngmi. Enjoy your locked bootloaders.
>>108672824Better than arm
>>108672844dickhead. (pic related)>>108672848worse than POWER7 at a minimum
Actual 16 bit x86 is dead now Windows is 64 bit only.
>>108672841No, but you can use it from a thin client. Everything will be done server side in the future. The concept of personal computing will be a little known topic for historians.
>>108668645So good, in fact, that they had to send the 8086 plans to Isreal and they cut the i/o bus in *HALF* there to cripple it (in Haifa… which must mean ‘half’ in hebrew) Then named it 8088 (americans thought it was better because bigger number)You can’t even make this shit up.Thank god I don’t have a pager anymore.
>>108672952> use it from a thin client. Everything will be done server sideLamo, you can do that now with VNC/MSTSC Even locally it’s painful. Doesn’t support digitizers properly either.Remotely to another city in your state it’s unusable.Don’t worry, when we break the speed of light on network communication then it will be fixed. In fact, the compile will be finished before we even started it!
>>108673012Speed of light is not a factor. Modern user interfaces are far more laggy than network ping. Normies are used to everything running like treacle. Normies don't even mind when modern games are running in the hundreds of ms of latency.
>>108668642Anon, gaming while on batteries is a strong sign of mental retardation.
>>108668605Gaming is the only thing that wasn't completely rigged by google. Chrome (by google) && Youtube (running in Chrome (by google)).The massive cope they must be going through is probably so debasing and demoralizing to them that they're at their last gasps...
>>108673214Mac is clearly the gamer's choice. Which begs the question: why is Apple not selling RGB gamer Macbooks as gaming laptops?
>>108668345You're correct, but you're more correct than you think you are.Since the pentium pro there is no such thing as "x86 CPUs" anymore.It's all random ass RISC CPUS with a translation layer.
>>108670080The champion in this category are the raspis and clones.
>>108672817>God please don't compare x86, the best cpu architectureOnly the 16-bit versions of x86 were good because there were actual reasons for doing it that way (read the manuals to find out why). The 386 and later were a Frankenstein slop architecture. Segmentation was supposed to be a feature that made x86 better than other architectures (and it has a lot of advantages in Multics) but the 386 didn't extend it right so people avoided it, and then 64-bit mode crippled it even more.
>>108673499What exactly do you think is wrong with i386? We are all in agreement that amd64 is an abomination, but i386 is good as far as I understand it.
>>108672635>arm is the futureAnd yet it needs to do all kinds of fuckery to make shit run thanks to it's genius design of not having any instructions on it while x86 just werks because it has them on the chip itself.Again, the only reason Arm has been pushed so hard was because it was more power efficient than x86, and both Intel and AMD have been hard at work to show that the only reason that was the case was not because of the architecture itself, but because of the chip design. These more "power efficient" Arm chips in the enterprise world are also less powerful than x86, so expect upcoming Xeons and Epycs that'll deliver more compute power at the same power draw as Arm chips.That eunuch and his bong pals came up with that architecture not long after Intel came up with x86, and failed to compete with it for decades until now? They're losers that got salty they lost when it mattered and are trying to force their way in. Same with electric cars, or Rust or Wayland.>>108672952There is a very good reason Microsoft was forced to release LTSC when they wanted to push everyone on the whole rolling SaaS format, and will be the reason why the whole cloud computing concept will forever remain a pipe dream. Your Internet connection goes poof, you're now liable for people's shit getting fucked up. Having a steady supply of AC is more reliable than the entire Internet infrastructure.Your cute doomerist stories are exactly what corposcum would want people to believe, but anyone with a brain, including these corposcum know it's not feasible and we ran code on physical machines for a reason. Again, Microsoft got their hands forced by their enterprise clients to give them an unshittified version of Win10 because mission critical shit matters and should never be tied to a fucking cloud. There'll always be a way.
>>108674357>steady supply of AC is more reliable than the entire Internet infrastructureThe internet was designed to withstand a nuclear attack on the continental united states
>>108674357> thanks to it's genius design of not having any instructions on it What nonsense are you on about. Modern ARM has a rich instruction set.
>>108674592>The internet was designed to withstand a nuclear attack on the continental united statesYeah no shit, ARPANET was designed for continental US and continental US only. Then very quickly grew into something much bigger and convoluted, including a variety of sea cables that get severed by sea life and more complex protocols and devices, not to mention the modern fragile architecture and general incompetence where colossuses like AWS, Azure and Cloudflare took down half the web with their downtimes. And you want to tell me that ARPANET being designed in the Cold War for continental US only to withstand a nuclear attack is the reason your flowery worded doom stories of "personal computing was a mistake that is now being corrected" are valid and can be done without any issues?All you're proving by bringing this up in relation to how the plan is for all computers worldwide to connect to remote servers in Silicon Valley to kill personal computing is that there's no point to replying to (You) since all you can do is make disingenuous arguments that make zero sense to feel like you're more smart than you think you are. And if you wanted to get a reaction out of me, hope this wall of text has tickled your willy. Not like it takes me much effort to type these out anyways.
>>108668246why did they call it x86 and not x78
>>108675413ikr they should've called it x67
>>108668645x86 was a toy cpu to run single user dos programs, while all serious workstations and servers were risc-based unix machines
>>108672635most software is actually written in high level languages with no dependence on a specific cpu architecture. issue is a lot of software is closed source so you only have access to a compiled binary version of it and that is what is cpu-specific
>>108675478And then it turned out the 5150 was a success because it came with enough documentation for anyone to become a programmer, and IBM's mistake of making it so off-the-shelf ended up becoming the biggest strength of the platform as it allowed for open development of the hardware, where x86 machines very quickly thwarted all those RISC-based Unix workstations. Especially once 3dfx started selling affordable 3D accelerators for x86 machines while SGI insisted on selling overpriced walled gardens, which ended up doing them in. Then of course Linux started thwarting Unix thanks to it being open, which was designed to run on x86 systems from the beginning.The 5150 and the x86 might've been "toys" in the 80's, but by 90's it became crystal clear that walled garden solutions such as ones SGI was offering simply couldn't compete with open platforms. IBM might've lost to the PC-compatibles, but they've planted the seed that revolutionized computing worldwide.
>>108675478initially, but then the 386 came out. flat 32bit memory space (functionally), mmu/virtual memory support, etc. while still slow and cheap, with fpu being a separate optional chip, it was enough to call it a serious cpu.now obviously most consumers just ran dos and treated it like a faster 16bit cpu, but consumers didn't need any more at the time. things like 386BSD at least demonstrated that it could run a serious OS
>>108675478>while all serious workstations and servers were risc-based unix machines"RISC-based Unix machines" were also toys. VAX and VMS weren't RISC or UNIX. Mainframes weren't RISC or UNIX.
>>108676100VAX got replaced by Alpha
>>108668246>x86 le good (no z800 or mk68)>UNIX le good (no VMS or NT)>C le good (no Pascal)>JAVA le goodDon't you see a pattern here?
