Max, when we were created by a loving being the angels whispered in our young ears the complete physics, chemistry, languages, anatomy, physiology and zoology and then for some reason, touched our philtrum to make us forget. Paradise lost. Some hacker once said, "Time is my weakness". Time. Got time to learn code? Got time to fix a pc problem that no one else can? Got time to learn all there is about the pc and software?
exdeath
Lifer
- Jan 29, 2004
- 13,679
- 10
- 81
Understand it? Yes. Have the infrastructure and ability to do it all myself from scratch? No. Not in several decades.
From the solid state physics of field effect transistors and logic gates, to the lithographic masking and etching processes to create an ASIC CPU, to the low level programming and instruction decoding involving counters, multiplexors, etc, to the storing of bytes in video memory interpreted by a DAC/differential digital transmitter and turned into a pixel on a screen in a CRT/TFT.
It's not too difficult to learn how a CPU works by writing your own in Verilog or VHDL and synthesizing it in a FPGA, or even building your own 6502 using a bunch of breadboards and 7400/4000 series discrete logic ICs. Start with the basics by learning how a carry look ahead adder works and work your way up to a superscalar out of order micro-op scheduler complete with virtual memory and translation look-aside buffers, cache, and process protection and privilege levels. And not just knowing how they work, but actually being able to implement the complex hierarchy of stall mechanisms and signals in hardware that make it all work and draw up a working transistor level diagram without having to look anything up. I can't even do that.
Of course if you can design the CPU and it's instruction matrix, programming it with the instruction set you designed becomes trivial.
There really isn't anything complex about a CPU if you start with the basics and work your way up and realize that each thing is result of a necessary evolution from the previous thing. If you try to start with a i7 your brain will explode. Start with a 6502 and go from there like the rest of the industry did.
And I'm just talking CPUs. Not high speed external buses, PCB design, DRAM design, storage systems, GPUs, peripherals, etc.
And little things we take for granted. For example, starting from scratch, with no other computers, how do you program the ROM by hand that a CPU needs on power on to start up? Now you have to build yourself a PROM burner on top of it, but how do you get the millions of bytes of startup and init code and BIOS level device drivers (eg HDD) to it without an existing computer in which to store it as a file?
Oh and that HDD is another computer in and of itself with it's own microprocessor, instruction set, cache, and closed loop servo feedback system... not to mention the exotic materials and precision manufacturing processes that go into the GMR materials used in the disc emulsion and heads...
And what about tooling? Building the tools to build the thing you want is an entire industry in and of itself with it's own set of issues. How you you build a precision milling machine without already having a precision milling machine to build it's parts with? How about the optics used in lithography that can focus fine detail in nanometers without being blurry or having impurities or imperfections in the glass? Optics is yet another field all to itself. And as mentioned the mining of materials, many of which do not just occur in pure usable form in nature and must be chemically isolated, etc. Something simple like aluminum even. You don't just pick up aluminum rocks off the ground and start smelting them.
Yeah. No way in hell is one person supposed to be able to do all this in one lifetime. I don't care if you have a IQ of 9000 and know it all, it's still not happening, there just simply isn't enough time in the the span of a single human lifetime for it to be physically possible. Knowing and doing are two different things.
Understand how it all works?
Yes of course I do.
Does that mean anything useful? Nope.
Those who know it all or almost all will be the first to admit they can't do it all.
From the solid state physics of field effect transistors and logic gates, to the lithographic masking and etching processes to create an ASIC CPU, to the low level programming and instruction decoding involving counters, multiplexors, etc, to the storing of bytes in video memory interpreted by a DAC/differential digital transmitter and turned into a pixel on a screen in a CRT/TFT.
It's not too difficult to learn how a CPU works by writing your own in Verilog or VHDL and synthesizing it in a FPGA, or even building your own 6502 using a bunch of breadboards and 7400/4000 series discrete logic ICs. Start with the basics by learning how a carry look ahead adder works and work your way up to a superscalar out of order micro-op scheduler complete with virtual memory and translation look-aside buffers, cache, and process protection and privilege levels. And not just knowing how they work, but actually being able to implement the complex hierarchy of stall mechanisms and signals in hardware that make it all work and draw up a working transistor level diagram without having to look anything up. I can't even do that.
Of course if you can design the CPU and it's instruction matrix, programming it with the instruction set you designed becomes trivial.
There really isn't anything complex about a CPU if you start with the basics and work your way up and realize that each thing is result of a necessary evolution from the previous thing. If you try to start with a i7 your brain will explode. Start with a 6502 and go from there like the rest of the industry did.
And I'm just talking CPUs. Not high speed external buses, PCB design, DRAM design, storage systems, GPUs, peripherals, etc.
And little things we take for granted. For example, starting from scratch, with no other computers, how do you program the ROM by hand that a CPU needs on power on to start up? Now you have to build yourself a PROM burner on top of it, but how do you get the millions of bytes of startup and init code and BIOS level device drivers (eg HDD) to it without an existing computer in which to store it as a file?
Oh and that HDD is another computer in and of itself with it's own microprocessor, instruction set, cache, and closed loop servo feedback system... not to mention the exotic materials and precision manufacturing processes that go into the GMR materials used in the disc emulsion and heads...
And what about tooling? Building the tools to build the thing you want is an entire industry in and of itself with it's own set of issues. How you you build a precision milling machine without already having a precision milling machine to build it's parts with? How about the optics used in lithography that can focus fine detail in nanometers without being blurry or having impurities or imperfections in the glass? Optics is yet another field all to itself. And as mentioned the mining of materials, many of which do not just occur in pure usable form in nature and must be chemically isolated, etc. Something simple like aluminum even. You don't just pick up aluminum rocks off the ground and start smelting them.
Yeah. No way in hell is one person supposed to be able to do all this in one lifetime. I don't care if you have a IQ of 9000 and know it all, it's still not happening, there just simply isn't enough time in the the span of a single human lifetime for it to be physically possible. Knowing and doing are two different things.
Understand how it all works?
Yes of course I do.
Does that mean anything useful? Nope.
Those who know it all or almost all will be the first to admit they can't do it all.
Last edited:
007ELmO
Platinum Member
- Dec 29, 2005
- 2,046
- 36
- 101
jacktesterson
Diamond Member
- Sep 28, 2001
- 5,493
- 3
- 81
Cogman
Lifer
- Sep 19, 2000
- 10,284
- 138
- 106
Tell me, exactly, the handshake protocol that goes on when you connect a USB3 device to a computer (without googling it).
This thread needs much more math.
I'm not going to go into the details. I've moved too far into the software world to remember the partial differential equations, multi-variable calculus, Taylor's theory expansions, gaussian surfaces, riemann surfaces, Laplacian transforms, Fourier transforms, (I wonder how many I've forgot), and how they work and why.
But you need them, to prove that your theories are correct, to be able to simulate systems instead of build them and watch them fail, and to be able to torture students
I'm not going to go into the details. I've moved too far into the software world to remember the partial differential equations, multi-variable calculus, Taylor's theory expansions, gaussian surfaces, riemann surfaces, Laplacian transforms, Fourier transforms, (I wonder how many I've forgot), and how they work and why.
But you need them, to prove that your theories are correct, to be able to simulate systems instead of build them and watch them fail, and to be able to torture students
Last edited:
SOFTengCOMPelec
Platinum Member
- May 9, 2013
- 2,417
- 75
- 91
Magic creatures called :-
Ultra
Small
Beings
The 3rd
U.S.B.3. for short.
Run about and collect the data, with their small hands (that's why the first part is called handshaking).
Apparently they can't stand the smell of the initial batch of haswell processors, which is why USB3 doesn't work well on the first chipset release.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 21K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter