I Designed My Own 16-bit CPU 

Hmmm 7 minutes beforr video was uploaded

@@offsetmonkey5382 sus

Discord server is very good everyone should instantly join

Very Nice

Why don't you team up with @Sam Zeloof to meke a homemade computer, he made an 1000 transistor silicon based chip.

Thank you! Also, think the biggest problem with this project is time. I believe anyone can do it, easily, given there are so many resources online. But finding time to do the research, and make the design can be difficult. I didn't know anything about computer hardware, but I did research (and watched a lot of Ben Eater), and then a month later I made this cpu. It is relatively basic, but it is a start.

Wow, are you just super smart or something @@AstroSamDev

​@AstroSamDev Even still, it takes incredible dedication to complete. Even if there's enjoyment to it, what you've done is beyond impressive!

Do you use programming language that were not made for Browsers ? (aka, not JS) If you answer yes to that, you're not a fraud.

​@@monad_tcpyou wouldn't be able to write this comment without JS

Broo this is not 4chan

@@le9038 just assumed it was

and after create own digital world where is made computer those emulate metaverse

@@le9038 ew... reddit

@@le9038 I think you'll find it's quite literally the opposite lmaoo

16 bit in 1977? That's insane

@@helloworld145 Not for a mainframe. But the price tag sure was insane.

Cool

@@helloworld145 There were 64-bit computers in the early 1960s.

@@herrbonk3635 64 bit computers must be crazy expensive in that year lol.

I can't imagine what it must be like for the people who were there from around the beginning of all this craze to where we are now. At some point I imagine you knew everything there was to know about but as time moved on new products, programs and such became available and after a while it just becomes too much to all learn and you just watch in awe at how dense and complex its become

@@handlehagglerthe lady who help invent voip is still alive…she was one of the hallmark individuals in the creation of the World Wide Web. It’s so wild to me.

I can do it in one

@@dimi5862 go on show me, everything from scratch, from nothing. As of this comment the date is Saturday 23rd july 2022. You can work during live streams to make sure you didnt cheat

i can do it in one (millenium).

@@maindepth8830 lol but true

@@dimi5862 prove it

Just read textbooks

These things are taught in engineering.

@@elakstein no they dont!

@@greatbullet7372 ok. I will correct myself. I learnt these things in college. Obviously not as deep as he showed in this video.

@@elakstein true... but we somehow dont understand or learn that but a person like this can teach it in just hours...

You think you feel under skilled I'm 22 and just learning visual studio code and Django

Finally something I can relate to

At least you’re not in denial. Less competition for us folks!

​@@0x_nietoh "folks"

Womp womp

Noooo. You just haven't gone deep enough to understand. Not understanding YET is totally different from being stupid. :)

Hello can u help me? Im new to coding can u link me to any good sources for learning python

@@memeology4138 learn c first. Python abstract just too much. First language should be c if you want to understand things. I would recommend "Let us c" for leaning c. We used this to learn c in college, then I learnt Java fir understanding object oriented language, then python.

@@memeology4138 im new too but im doing c and js

@@elakstein thx for ur help

@@abcdefghjijklfgu why c and js? learn one language first, master it. plus, js is usually nothing without html/css, unless you're learning react and stuff.

I was planning on doing just that, it might just take some time though. I also want to do a design that doesn't use breadboards, and instead have a few custom PCBs printed.

@@avinadadmendez4019 Cool, I have never heard of that program, I'll look into it more. Thanks!

@@AstroSamDev I would buy it if it's not too expensive! I really like this project and i think it would be really fun to develop own programms that run on real hardware!

@@AstroSamDev @AstroSam You can also put your design on an FPGA, I don't think I have ever seen one done on youtube before for a homebrew cpu. For those who don't know; FPGAs are circuits that you can load any digital circuit on it. For example you can put your circuit on an FPGA and map the input output pins of the top module and the circuit is now the circuit you created.

@@cemalettincembelentepe8943 FPGAs have been used for soft CPUs for a very long time now (well over a decade), it would be more interesting to see a hard CPU manufacturing approach and it would be unique for these homebrew CPUs unlike using an FPGA. Granted he'd be stuck getting them as samples on a really old process node (>=28nm), but that's more than good enough for this purpose

I understood about 0000000000000000% of it. Wondering WHY he did it, When there’s stuff already available and cheap these days, I guess it boils down to “Because I can”

It's just TempleOS minus the schizophrenia

@@BigOrse so its just boring then

@@test-zg4hv What so I should praise the man for being completely insane?

@@BigOrse I'll wait on that diagnosis..

Not coming here to defend myself, I don't care if what I said was stupid. To anyone saying I'm disrespectful though I suggest you maybe actually look into Terry Davis as a person, because he was a lot more disrespectful than I am. Anyway RIP Terry. I may not agree with his ideology but can't deny the man was a pretty legendary programmer.

256 addresses and some external logic should be enough to implement a microcode architecture.

A-level??? Dude that's unreal. I built an sql table viewer in VB...

Fixed instruction length isn't that bad, it's what RISC uses.

@@shinyhappyrem8728 considering the memory limitations, every instruction being 3 bytes long is an issue (command operand1 operand2), some instructions didn't need operands, such as halt, but would still need 3 bytes in RAM.

@Ramtin Javanmardi If you understand how the architecture works, writing an emulator isnt really too hard, you could even write a simulator if you wanted, but you need to get instruction timings and all sorts of hardware stuff simulated too, which can be a hassle. Basically, you need really 2 core things for an emulator, an ISA, which you should have if you designed the microArch, and some way to interpret the ISA, really this will probably be high level logic in your chosen programming language. I simply used a case switch on binary opcodes that called methods with the operands (remember, because this is an emulator, memory read and write timings dont matter), but a better approach (and the one I used on my Z80 simulator) would be to use instruction classes and loop through an array of these classes (well, cast to interfaces) testing if the opcode can execute a given function in the class. How you handle IO is up to you, I didn't, but memory mapped IO is pretty easy to implement and can be done by simply accessing the RAM object of your emulator. My Z80 simulator is open source, and was written for a university project, so it might help, it can be found here: github.com/Lewinator56/z80Sim (you will need .net core 3.1) Writing an assembler is kind of different, in reality at the easiest level you are simply going to have a defined assembly language and just convert it into opcodes, which is pretty easy to do. If you want to include functions and variables (by name) in the ASM then you need to process symbols too. I simply stuck with a conversion from ASM to binary.

alternatively you can also be lazy like me and simply build your circuits in a logic simulator like "Digital" (it's like Logisim but faster and with extra features) and then just export them as Verilog/VDHL. it works surprisingly well actually.

yes this

And at this point I would have thought that chips from Intel and AMD are pure evolution .. no revolutionary steps. Steps are not agile. Maybe they include experimental modules in every chip, test them, and if it works, you get the next iteration ( more expensive ). An FPGA and a chip from Intel have nothing in common. Intel chips are optimized down to the photolithography layout and full analog simulation (spice) of the transistors. FPGA does not give you this. The days of the Pentium DIV bug, where a professor from a university uses an ad hoc script to get his algorithm from university software to some proprietary CAD at intel, are long gone.

@@ArneChristianRosenfeldt The advantage of using an FGPA instead of individual chips for this project is, that others could make use of it too by simply reprogramming their FPGA. There is already a real CPU design called Magic-1 from Bill Buzbee. The problem is, others have to recreate it, if they want have a copy of it. Search for Magic-1 on YT, there are some videos about this impressive project.

​@@ArneChristianRosenfeldtwouldn't FPGA be a bit more like a drawing board at that level?

is it linux based?

@@WTBuilder _"Interjection!!"_ - Stallman Ace Attorney

I built the entire observable universe by scratch with real breathing life forms in it.

👍

Yeah. I've wanted to build my own computer for decades, but my brain just goes, "what are these facts NOPE!" lol but it's getting better slowly.

I did that last year too! I don't know what software you used, but here we did it with diglog and it was a lot of fun ^^

@@mathismartin3092 Our exams were mainly done on paper, but we used logisim for practicing! Of course we stuck to no more than 4 or 5 bits for the instruction set, and mostly doing just read, write and math instructions, but it was really interesting to figure out how to do certain things :D

Probably, that text editor is just begging to be transformed into an interactive command line.

@@AstroSamDev AstroOS coming :000

@@AstroSamDev rewrite the Linux kernel in Z# and make z#/Linux /s of course

Lol aside, that's actually not impossible: see Collapse OS for how to bootstrap from almost nothing to working OS.

Agree. My university had us build a single cycle 8-bit MIPS processor on a Xilinx FPGA. A very simple processor, yet probably the best project we've done so far.

if it supports color for text, 1 character is enough: you simply swap background and foreground color which turn ▀ turns into ▄ .

or a way to write arbitrary data into unused characters and then display that with the faster character rendering

This would make the drawing routine more complex and thus slower. Better design a bit blitter and let the bit blitter hardware do the rest.

Don't forget he also wrote a port of Pong to run on said computer...

Bro this is like jdh shit right here

*straight

And he's ONLY 16. When I was 16, I was barely writing hello world in C++.

​@@PiyushBhakatBits per age. This man will only grow stronger in time and consume more bits

how many habibibytes is that

@@axmoylotl 2 hamoodibytes

I'm never going to use that childish sounding garbage. 1024 bytes is a kilobyte and SI has no business even being included in the discussion, I don't even care that they're using the same letters for prefixing, it's still going to be powers of two. Hard drive manufacturers have caused and are continuing to cause harm to the industry by being such cheats and now we've got douche-nozzles who want to redefine all of our terms instead of taking the manufacturers to task. It's almost as bad as changing BC/AD to BCE/CE because they don't like the implications of how the names originated. Stop redefining terms.

Nobody cares about kibi, kilobyte is 1024 in my heart.

@@tissuepaper9962 same here

Was going to mention this. It would make timing quite a bit simpler, no more arbitrary wait N cycle loops, just execute game logic once on every vertical sync interrupt. :)

and Logisim's speed does depend on whether or not one is using the default 'chips' inside it or not, it has math operators and register modules, and whatnot, but if you eschew that and go down to the individual gates in sub-circuits....

He does say that logisim is simulating every logic gate, and that's why it's slow. Seemed sufficient to me.

Yeah I was a bit confused at that. Why would they differ in speed so much when they’re both emulators?

@@Costelad Imagine instead of executing an add instruction directly, you simulate the propagation of electricity through each part of the circuit implementing an add operation.

Intel x86 is a little more complex than making a 16 bit cpu.

You want Sam Zeloof for that :P ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-s1MCi7FliVY.html