Hi /g/
I'm an average programmer (C/C++) and I work as a sysadmin. I don't intend to do ComSci or electrical engineering but i'm interested in how computers actually work. By that, I don't mean knowing LOL LE CPU AND RAM xDD, I mean actually gaining a ground up approach, from electrons to logic gates to ISA
Are there any good books for someone like me to learn about this sort of stuff without doing a degree?
Cheers
Yes.
>>51649165
install gentoo
>>51649173
>Yes.
Such as? Everything I've read so far stops at levels of abstraction to which I want to go deeper, such as saying "it is handled in the ALU" well I want to know what is IN the ALU and how it works
E.E.
Read this book.
Check out nand2tetris
The more general term is 'computer architecture and organization'
>>51649165
Computation Structures is the end-all beat-all classic on this subject.
>>51649165
The Elements of Computing Systems: Building a Modern Computer from First Principles (aka From NAND to Tetris).
It's made by kikes but it's good. Don't buy it though, it's available for free on libgen.
![61rr4idpyzL._SX430_BO1,204,203,200_[1].jpg](https://i.imgur.com/54VbIri.jpg "61rr4idpyzL._SX430_BO1,204,203,200_[1].jpg")
>>51649165
Roth Fundamentals of Logic Design
Structured Computer Organization by Tanenbaum
or the book I'm reading right now:
Computer Organization And Design by Petterson & Hennesy
>>51649726
P&H is a load of wank.
>>51649212
Step 1: read a book on analog circuits to learn how transistors work
Step 2: read a book on cmos vlsi
Okay, first you need to read through the chapters relating to electrostatics and electromagnetism in any version of Halliday/Resnick's Fundamentals of Physics. Beyond that you need to read a book on intro to electronic engineering and digital system design. From there begin reading up on commonly used circuits and the set up in systems you find interesting.
You won't get a flawless and insanely deep understanding of it all because you basically have to major in computer or electrical engineering to get there, but you'll have a definite working knowledge. Also, don't start too low because that would mean teaching yourself quantum mechanics (though rudimentary wave behavior and quantum mechanics are covered in Fundamentals of Physics, last set of chapters).
You're not just jumping into the deep end to learn how to swim, you're announcing that you're jumping in and going to swim the English channel.
Get an Mx6 board and start with the Freescale stuff, and gradually work your way down to the transitor level.
>>51649835
I don't personally see anything wrong with learning electromagnetic physics to better understand what's happening in circuitry before getting into the design aspect.
>>51649749
Somebody's upset that they made a C in computer architecture
Come study CS to Wroclaw Uni of Tech they teach you that as an obligatory Course
>>51649165
The Art of Electronics by Horowitz is great for some general electronics knowledge.
As another anon recommended, C O D E is probably the great compendium for technical computer knowledge.
>>51649939
You know me too well, anon.
>>51649767
Computation Structures starts with transistors and how the various logic families are implemented, and ends with writing a multitasking OS on top of one of the four microprocessors built from scratch in the book.
Complete schematics.
>>51649376
Seconding this
>>51649930
fucking this!
For the even lower level parts try http://www.amazon.com/Logic-Computer-Design-Fundamentals-Edition/dp/0133760634/ref=dp_ob_title_bk
>>51649165
>download garry's mod
>download wiremod
that's how i learned logic desu
>>51649165
The average programmer doesn't know how a computer works? No wonder you're losing work to Sundheep.
>>51651813
The average programmer probably can't tie his own shoes.
>>51649930
>that book cover
y tho?
>>51649173
/thread
It's not going to helpful or probably even all that interesting. If you don't use all the knowledge you learn by applying it to problems, you'll probably just forget it quickly anyway even though you "think" you learned it. Shit is complicated and relatively esoteric as far as writing code goes.
If you understand branch prediction and cache locality, I'd probably just call it day.
Literally starts from as low as you can go: a single switch.
http://download.savannah.gnu.org/releases/pgubook/
Plus: open source book.
>>51653820
My bad, that's the wrong book.
The Elements of Computing Systems
https://mitpress.mit.edu/books/elements-computing-systems
Tldr
1.Semiconductors physics electrons and holes , n/n+/p wells to make transistors
2. Some transistors in a bunch of configurations make logic gates NAND, AND, OR,XOR , using logic levels for high and low.
3. A bunch of logic gates makeup boolean equations for discrete systems. Multiplexer, decoder, flipflop, etc.
4. A bunch of mux,/decoders/flip flops make finite state machines
5. Together with FSM and discrete systems you make a system like the ALU, Interrupt handler, fetch , decode , execute units and eventually a CPU.
6. RAM can be flip flop based or capacitor based.
7.Join your CPU , RAM, and other peripherals and you have a very basic computer.
Nowadays you use VHDL to make digital hardware. The next step is making code for your CPU
1. First in raw hex
2. Then you use another computer to program and assembler and using nmemonics to make programming easier.
3. You either use your assembler or a cross compiler to make a C compiler for your CPU.
4. Program the OS in C
5. Program the interpreter or other compilers
6. Program even more to make basic utilities.
7. Continue programming
>>51651920
as an average programmer I can confirm this
As a computer engineering student I like these books:
CMOS VLSI Design: A Circuits and Systems Perspective.
The Art of Electronics 3rd Edition.
Computer Architecture a Quantitative Approach.
>>51651813
And neither do you to detail.
>mfw first learning how superscalar processors actually work
It's the tightest shit.
>>51649656
Second. Currently doing the course for that in uni
>>51653959
Most would use Verilog instead of VHDL, but otherwise yeah.
>>51649376
I have this book. Is what he's looking for really in it? I haven't finished it but this would be a good incentive to pull it out and do so.
>>51649930
Heh, I still have this book from when I studied EE for half a year. Didn't make much sense to me.
Might take a look at it again now that I've got a CS degree.
>>51649165
Fundamentals of Digital Logic with VHDL
No introduction needed, aproaches several subjects, any eddition is good....
Plus, it allows you to adquire practical knowlege of VHDL ( not a programing language, but a hardware description language)
>>51649376
Thirding this. However the author never covers how control signals work which kind of bothered me. Since he said he would show how to build a computer entirely from scratch.
>>51656495
What is a good book to learn VHDL if I already know about digital logic?
>>51656763
The same book. They have a lot of example, exercices and pratics implementations. that you can use to learn about. I would link it to you, buit i only have a physical copy
>>51657140
...what are physical addresses.
>>51657140
Thanks
Read computer organisation and design. I have the 5th edition, and it goes into as much detail as you would want.
>>51649165
You need a solid background in circuits first. You have to start from the bottom now.
Source: I'm an EE. Read Sedra and Smith. Or ask me. I design CPUs.
>>51659434
>Sedra and Smith
5th edition is available online for free, if you don't want to pay for the 6th or 7th edtion.
Source: broke as fuck E.E. student
>>51649707
FUCK THIS BOOK
>>51659562
>Fundamentals of Logic Design
Why? Because it is bad for you?
>>51649187
kill ys
>>51659434
How do you solder CPUs?
I have this free electrical engineering course saved on my computer, if the thread's still up when I get home I'll link it for you.
>>51649391
This.
It will help you by guiding you through the process of building a simple computer from the logic gates up.
>>51649376
this.
