Brewing computers at home

4bit, 8bit, 16bit, 32bit, whatever bit, ternary etc. Single CPU as well as much more interesting and rare among homebrewers Multi CPU, homogenous, heterogenous and other ideas Schematics, ideas, software, electronics thread. https://hackaday.com/tag/homebrew-computer/ https://hackaday.io/list/2402-homebrew-computers Any brewers here?

>>5561 A homebrew computer has always been my "yeah I should do that someday" project. I can never decide how oldschool I want to go, so I never seriously put planning into it. Half of me wants to make a SBC with a 6502 or 8080 CPU and a modern MCU for i/o. The other half wants to fully commit to built from scratch, maybe even bootstrapped without an external computer. The zachtronics guys made an example of the latter extreme. http://www.zachtronics.com/gibson/

if I had the money I'd mess around with GreenArrays chips and similar setups.

A homebrew ring I found https://www.homebrewcpuring.org/ringhome.html >>5575 why not make both versions? >The zachtronics guys made an example of the latter extreme. http://www.zachtronics.com/gibson/ cool link, thanks >>5580 any precise ideas?

>>5586 >any precise ideas? I like the idea of making an interpreter specifically for the architecture. You could take advantage of the highly parallel architecture by doing something like CPU pipelining.

>>5592 So it would be multicore/multi CPU architecture?

>>5601 yes, that's the main point of the F18A chips themselves. They're fully asynchronous(they don't even have clocks) and designed specifically to be used in large arrays with fast I/O between them.

>>5602 It's 1st time I met these chips. Is that some novelty or just obscure stuff that has been around for a while? What's the perspective/usage for them in real life aplications?

>>5603 >Is that some novelty or just obscure stuff that has been around for a while? They've been making real existing chips since at 2011, and they still make new arrayFORTH releases. It's unclear if they are used by anybody but hobbyists. >What's the perspective/usage for them in real life aplications? the main advantages are that they are very fast (instructions take less than 6 nanoseconds to complete) and very low power. This means they have good heat characteristics and are probably more physically resilient for it. Probably the main use case is for embedded applications. They can also compete favorably with FPGAs when configured properly. However, it seems to me that the GA people have tried to make too many novel innovations, and that probably scares off corps.

>>5580 >>5592 >>5614 I like forth, but I always thought GreenArrays was a gimmicky piece of hardware. The architecture kinda reminds me of the videogame TIS-100. It's sold as low power, but significantly more expensive and complicated to use than other chips at the same power consumption, so I doubt they have many clients. Corps just want a microcontroller with a C toolchain and whatever peripherals they want easy to use so they can easily replace programmers for it. >>5586 The homebrew ring is neat. I'm surprised that there are multiple relay computers, but no discrete transistor computers.

Trying to implement SHA256 and RIPEMD160 using discrete components. This seems like the most secure way to make a cryptographic subprocessor for key generation, encryption and signing. Data input by diode matrix, output by integrated 7-segment display. Should be able to finally verify whether compiler backdoors exist.

>>5615 >It's sold as low power, but significantly more expensive and complicated to use than other chips at the same power consumption, Well it's low power but has pretty impressive speeds and parallelism for chips with similar power consumption. I think the thing is that it's targeting such a niche use-case: FPGAs can be faster and more parallel if you don't care about power consumption and microcontrollers are a lot cheaper if you don't care about speed, and of course nobody wants to pay to train their programmers in forth. They might have a few clients but I really doubt they get much business. It's really a shame because I do really like the design of the F18 chips. Asynchronous computing is just more fun to think about :P

Is there open source hardware capable of running Temple/Shrine OS?

>>5629 you would need an x86_64 CPU, which is licensed by Intel. So, no.

>>5630 *x86 is licensed by intel, the 64-bit extension is licensed by AMD

>>5630 Thanks for the quick reply. Figures, even if someone did manage to reproduce the architecture, shutting it down would be top priority. Porting an OS to another instruction set like RISC would likely require a second coming of TAD-caliber autism.

>>5630 >>5631 and Chinese clones are made w/o giving a fuarrrk about licenses from both LOL

>>5634 Republic of China obeys international laws, all (2 of them) Chinese x68 and AMD64 processor developers are located on Taiwan island. They have assembly facilities in occupied communist China, and a joint company to produce AMD64 processors for communists in Shanghai, but it still uses VIA Technologies architecture designs and patents.

>>5642 >Republic of China obeys international laws I checked my calendar if it's not April 1st.

>>5652 The Republic of China is not the same country as the People's Republic of China. Don't be a dumb fuarrrk.

>>5660 If you didn't mean People's Republic then replying to post >>5634 makes absolutely zero sense. Unless you have brain damage.

>>5642 Pretending Taiwan is real China is like pretending Miami is real Cuba. Don't let reddit infect you. Nb for off topic.

>>5561 Well, he had the right hat

Lads, I am actually making a CPU maybe not related but I will gift it to my professor. It's also really inefficent uses MMIX by Donald Knuth with a custom ASSembler. If anything like an entire computer comes out of it I will notify lol!

>>6505 This turn out well?

>>6487 >>you have to conform guys being eccentric and a maverick that makes a communist utopia online is cringe guys don't think differently guys

>>6995 * g0 y em to think the hat joke could have been just pruned but no course not

Anyone into analog computers? Solving complicated partial differential equations and complicated integrals in matters of miliseconds and power consumptions of microscoping scale is what makes me hard.

Cool homebrews https://www.homebrewcpuring.org/

Retrobrewcomputers is a good wiki/forum for all kinds of homebrew computers. https://retrobrewcomputers.org/doku.php

>>7301 Stuff like this does leave me erect. I'm big into analog logic circuits for synthesizers, just feeding different enveloping voltages and see what gates get spit out. You can set up pseudo random rhythms with creative "patch programming." Underneath all that I'm sure there is a lot of math theory involved. Both examples you show in the pictures give me design ideas.

>>10411 Have you heard about that analog thing? https://the-analog-thing.org/ there is a link to their websites

>>10497 This would be perfect for a Buchla, Serge, or other Banana cable equipped synthesizer. The only worry I would have using the computer would be outputting voltages that would overload the circuits. Then again you could DIY an attenuator for this purpose. Thanks for the information!

>>10506 your welcome, always happy to help

>>5575 I contacted Zach and he had this to say "It's very sold out, but considering that the computer never actually worked I don't know why you'd want any!" Somewhat disappointing but it is what it is.