Is it possible to opensource hardware? What technologic leaps will it require to make the jump?
Maybe with 3D printing? You'd just need to insert rare earth materials in it? Can that be done? No idea I'm not a 3D printer just thinking aloud.
risc-v is open source
>>107904622All you need to do is ask the Abrahamic owned company who doesn't actually make RAM to give you the loiscense to make the RAM.
>>107904622Why isnt there a service where I donate a percentage of my machine's compute, and in return, I get to use a burst of other people's compute when I need it? Like if I have a 32gb machine with a decent cpu and gpu and overnight I let the europoors run code on it and i accrue credits to do the same to their computer when I need it
>>107904622the truth is, you don't need more than late 90s hardware to meet all your computing needs. just jailbreak any modern phone and you're set.
>>107904985sure anonlet a random russian fag run code on your machine...what could go wrong ?
>>107905032What if the russian fag isn’t running code on your computer but only the company machines run code on your computer? Lets say i have to make a giant render: wouldn’t it be possible to subdivide it in 100 tasks and execute them in 100 different computers at the same time?
>>107905032Non-issue. VMs/sandboxing exists
>>107904622Fucking lines man just get a CAD program and a vacuum jar.
>>107905086a smart russian tranny can break out of these
>>107904622>What technologic leaps will it require to make the jump?silicon die fab house vending machines at every corner. send the design via some app, go back later to collect the chip
>>107904622Lots of hardware is open source
>>107905071>>107904985you could run a Flux nodehttps://runonflux.com/
>>107905126If they could they'd have crippled the entire cloud and VPS industry already
>>107905071Render farms exist and it's like you describeI don't know if something like your first post exists, for "general" computer usage
>>107905179yea but useless until we get some jlcpcb thing for dies lol
>>107904622That's called Communism
>>107904985the idea of sharing computing power is not new but funny enough if you play around with it and talk more about some of the implications it can lead to a 3 days ban although this was the case like 4 years ago and now i think it safe.the idea of sharing compute is bad because you can only share efficiently task completion: you cannot really share RAM, CPU or GPU because of latency. you can share HDD of course, but even HDD share has to be of non-vital for the OS parts. But sharing task completion is an old crypto idea and like most crypto projects it fall apart because 99% of the crypto companies were in it for the money. Lets take for example Mysterium - it used to be a very hot topic here as mentioning it was a ban reason. What you essentially could do it share your IP address via a raspberry pi note for somewhat good money. The problem of course was that 95% of the traffic was illegal as you are essentially providing residential IPs to random people around the world. Yes, people even here were constantly bragging about how they use it all the time to ban evade and how they can post ANYTHING here and get away with it. Of course this meant that if you provide IP addresses on the Mysterium network they will all be perma banned on 4chan but also the police will visit you. ..... But it was not the rampant illegal activity that made Mysterium close and become a normal VPN - it was just a company folded by an indian woman chasing quick bux during the crypto boom. Most of the "share computing power" with others companies all are / were in the crypto sphere and they all vanished by 2023. even the example given in >>107905190 with Flux instantly has "web3" on its page..... it will not last long.Only hard disk space sharing is viable. All other type of "compute sharing" is a meme. Also all types of "compute sharing" will be used for illegal shit. I hope you like the police because you will be seeing them.
>>107904780I am going to look into this more.
>>107904622>opensource hardware?i am not sure what that means to you. the home pc is the home pc due to being more of a standard that anyone can build to. the market has consolidated somewhat.
>>107905427Look, he is scared. Be very afraid of the future...
>>107905427It's closed sourced and being gated off from society where meeting demand is a risk for investment as well so that's not going to be the case in the near future. Things are changing.
>>107905427good luck releasing you x86 CPU.
>>107905388>local models>ci runners>distributed database nodes in some consistency models (ie. I promise to retain 100mb of your db in my ram for the next 15 minutes)Just off the top of my head. I agree there's limitations to what kind of software would be applicable but its not useless
>>107905449i don't know what any of this means in the context of my post personally, and i don't think you know what you're talking about. it looks like you're operating on your own ad hoc definitions that you made up for this thread and you haven't really explained with your esl skillsmaybe do a little reading. https://en.wikipedia.org/wiki/IBM_Personal_Computer>Powered by an x86-architecture Intel 8088 processor, the machine was based on open architecture and third-party peripherals.https://en.wikipedia.org/wiki/IBM_PC_compatibleAn IBM PC compatible is any personal computer that is hardware- and software-compatible with the IBM Personal Computer (IBM PC) and its subsequent models. Like the original IBM PC, an IBM PC–compatible computer uses an x86-based central processing unit, sourced either from Intel or a second source like AMD, Cyrix or other vendors such as Texas Instruments, Fujitsu, OKI, Mitsubishi or NEC and is capable of using interchangeable commodity hardware such as expansion cards. Initially such computers were referred to as PC clones, IBM clones or IBM PC clones, but the term "IBM PC compatible" is now a historical description only, as the vast majority of microcomputers produced since the 1990s are IBM compatible
>>107905502i really think that only task completion and storage can be shared all other type of sharing has to have some latency reduction mechanism like the distributed pi networks that use fast LAN communication.
>>107905495good luck releasing your any kind of cpu. you're not going to be throwing one together in the kitchen my man, x86 or any other architecture. easier to make your own car. a bunch of hardware is in some way, although maybe not how the op understands the term, open source or open spec. you're still going to need a fab or whatever manufactury to make it. so the ops question in that context would seem to be instead, will it be possible to make nanometer scale circuitry at home? not for a long fucking time if that's the case.
>>107905518You still need intel's permission to use the instruction set, so 86x opensourced is still fucked. Stop dancing around the monopoly and just make new forms of assembly.
>>107905548I don't think it's impossible, but I think communities will have to pool resources together to make it work as non profits that cannot sell to corporations.
>>107905569>Stop dancing around the monopolyi'm not dancing around shit homeboy. the op's question was is it possible to make 'hardware' open source. there was no specification about being solely cpus. the pc standard is an open architecture outside of the cpu and that was by design by ibm. >and just make new forms of assembly.well i'll fucking get right on that nigger
>>107905591>I don't think it's impossible,you don't think what's impossible
>bought 256 gb ddr5 and a 5090 because I'm not a luddite and knew how things would gofeels good
>>107905639welp, it was well worth it for that picrel, lmao
>>107904622You can open source the hardware, you can't ever reproduce a fab in your basement. I think the most hobbyists have ever done are 192KB chips with no error checking.
>>107904622If by "hardware" you mean chips, the architecture can be open sourced but the problem is turning that architecture into a real design that can be manufactured without breaking and then manufacturing it.Insane fab costs aside, TSMC and the other manufacturers keep the specs of their newer process nodes secret very tightly. I work at a chip design company and the only people who get access to the PDKs for 20nm and under processes are the people from teams that strictly need them.Some old processes at like 180nm are public, but you can't make competitive digital circuits with that. Making something like a CPU using that node is posible, but it would still cost a ton of money to manufacture it.TL;DR: It's too secretive and costly to manufacture to be open sourceable, even if it could be done. It's like asking whether ICE engines could be open sourced.
>>107905689Yeah, I think OP was thinking of ways to what it would take to get to that level, but everyone on this board is a professional at misinterpreting the most basic things.
>>107905840180nm is fine for me. I'll just have to put a few together, but it's fine.
>>107905865Yeah you'll have to pull a few together. A few tens of millions of dollars.
>>107906125That's like a small business owner level of financing.
>>107904847Muslims own oil anon, they are not in tech business
>>107905285It's called TSMC
