4 ms·

No need to posit a conspiracy here. People died, and there weren't enough new people entering the field to learn it all before the critical mass of knowledge d

by ken 7y ago

No need to posit a conspiracy here. People died, and there weren't enough new people entering the field to learn it all before the critical mass of knowledge died with them. I believe that the bulk of knowledge, in most fields, is passed from person to person through active practice, not through books. Books, as a cheap way to widely distribute knowledge, are relatively recent innovation, and most of what I know about any topic is not written in books. Or if it is, it's almost an afterthought, not the primary method of knowledge transfer. What book did you read to learn to tie your shoes, or drive a standard transmission? On a related note: what is humanity's Bus Factor for creating a CPU? I know a little about semiconductors from school but I certainly wouldn't know where to start, even for a tiny 8-bit one. I doubt Intel writes down everything they've figured out. Could 2 or 3 especially unlucky "city-killer" asteroids send us back to a pre-information-age society?

person_of_color 7y ago
Relevant book: A Canticle For Leibowitz.
lloeki 7y ago
What’s your goal? Rebuilding an Intel class modern CPU or just making a crude one? Transistors give you logic gates and opamps, from which you can build computation and memory, add a clock to apply input from a certain address (the PC) to a set of logic gates (the instruction set) that can read and store and you have a very crude (and very inefficient) programmable CPU in no time (compared to something even remotely like a 8080 or even a 6502) A good approach to get more knowledgeable is going bottom-up from transistors and top-down from FPGA (and suffer Verilog/VHDL), attempting to experiment and design your own instruction set while trying to understand the physics and logic, progressively bridging the gap in between in both directions.
TheOtherDave 7y ago
If you’re talking about losing the ability to recreate any particular relevant CPU, then maybe. But computers in general? No. I suppose we could be exceptionally unlikely and lose access to all the modern fabs, but the knowledge of CPU design is too widespread to for us to plausibly lose the ability to make new CPUs altogether.
axaxs 7y ago
This seems to happen a lot. One people seem to keep chasing is Damascus steel. Even in more recent history, too. I like to research fine weapons. Apparently the process Colt used on their pistols to color and shine them is completely undocumented and now unknown. Similarly, it's said there are not enough people alive that know how to handfit a revolver. I'm not sure how true the latter is, but really makes you wish Wiki or the internet existed further back.
- dboreham 7y ago
  Also, supposedly, the styrofoam used inside thermonuclear weapons.
tzs 7y ago
A whole lot of people know how to make a CPU. It's not as hard as you probably think it is. What's hard is making them as complex, fast, and small as we do nowadays. But if you can get by with something not as complex or fast, and that takes up most of your desk, thousands of people know how to do it. Got 120 hours and want to become one of them? Take these two courses from EdX, which are 60 hours each: "Computation Structures - Part 1: Digital Circuits" [1] "Computation Structures 2: Computer Architecture" [2] The first teaches "[...] digital encoding of information, principles of digital signaling; combinational and sequential logic, implementation in CMOS, useful canonical forms, synthesis; latency, throughput and pipelining". In the homework and labs you design and implement (in a simulator) a 32-bit ALU. The second covers "[...] instruction set architectures and assembly language, stacks and procedures, 32-bit computer architecture, the memory hierarchy, and caches". In the homework and labs you design and implement (in a simulator) at the gate level a 32-bit RISC CPU, except for memory. Memory for registers and program is given as a black box--by this point you know enough to design that, but it would just add a lot of components and complexity, and slow down the simulation, and the time spent dealing with it would distract from learning the topic of this part of the course. That would fit better with the first part of the course. I've taken these, and can say they do a good job of teaching what they say they teach. There's also a third course in the series: "Computation Structures 3: Computer Organization" [3] That covers "[...] pipelined computers, virtual memories, implementation of a simple time-sharing operating system, interrupts and real-time, and techniques for parallel processing". In the homework and labs for that one, you optimize your CPU from the second part for size and speed, and make it support time sharing operating systems. I've not taken this one. [1] https://www.edx.org/course/computation-structures-part-1-digital-circuits-3 https://www.edx.org/course/computation-structures-part-1-dig... [2] https://www.edx.org/course/computation-structures-2-computer-architecture https://www.edx.org/course/computation-structures-2-computer... [3] https://www.edx.org/course/computation-structures-3-computer-organization-2 https://www.edx.org/course/computation-structures-3-computer...
- acqq 7y ago
  It’s not the logic but the technology that is not widely known: the new fabs cost billions.

by ken 7y ago

person_of_color 7y ago

Relevant book: A Canticle For Leibowitz.

lloeki 7y ago

What’s your goal? Rebuilding an Intel class modern CPU or just making a crude one? Transistors give you logic gates and opamps, from which you can build computation and memory, add a clock to apply input from a certain address (the PC) to a set of logic gates (the instruction set) that can read and store and you have a very crude (and very inefficient) programmable CPU in no time (compared to something even remotely like a 8080 or even a 6502) A good approach to get more knowledgeable is going bottom-up from transistors and top-down from FPGA (and suffer Verilog/VHDL), attempting to experiment and design your own instruction set while trying to understand the physics and logic, progressively bridging the gap in between in both directions.

TheOtherDave 7y ago

If you’re talking about losing the ability to recreate any particular relevant CPU, then maybe. But computers in general? No. I suppose we could be exceptionally unlikely and lose access to all the modern fabs, but the knowledge of CPU design is too widespread to for us to plausibly lose the ability to make new CPUs altogether.

axaxs 7y ago

This seems to happen a lot. One people seem to keep chasing is Damascus steel. Even in more recent history, too. I like to research fine weapons. Apparently the process Colt used on their pistols to color and shine them is completely undocumented and now unknown. Similarly, it's said there are not enough people alive that know how to handfit a revolver. I'm not sure how true the latter is, but really makes you wish Wiki or the internet existed further back.

dboreham 7y ago
Also, supposedly, the styrofoam used inside thermonuclear weapons.

tzs 7y ago

A whole lot of people know how to make a CPU. It's not as hard as you probably think it is. What's hard is making them as complex, fast, and small as we do nowadays. But if you can get by with something not as complex or fast, and that takes up most of your desk, thousands of people know how to do it. Got 120 hours and want to become one of them? Take these two courses from EdX, which are 60 hours each: "Computation Structures - Part 1: Digital Circuits" [1] "Computation Structures 2: Computer Architecture" [2] The first teaches "[...] digital encoding of information, principles of digital signaling; combinational and sequential logic, implementation in CMOS, useful canonical forms, synthesis; latency, throughput and pipelining". In the homework and labs you design and implement (in a simulator) a 32-bit ALU. The second covers "[...] instruction set architectures and assembly language, stacks and procedures, 32-bit computer architecture, the memory hierarchy, and caches". In the homework and labs you design and implement (in a simulator) at the gate level a 32-bit RISC CPU, except for memory. Memory for registers and program is given as a black box--by this point you know enough to design that, but it would just add a lot of components and complexity, and slow down the simulation, and the time spent dealing with it would distract from learning the topic of this part of the course. That would fit better with the first part of the course. I've taken these, and can say they do a good job of teaching what they say they teach. There's also a third course in the series: "Computation Structures 3: Computer Organization" [3] That covers "[...] pipelined computers, virtual memories, implementation of a simple time-sharing operating system, interrupts and real-time, and techniques for parallel processing". In the homework and labs for that one, you optimize your CPU from the second part for size and speed, and make it support time sharing operating systems. I've not taken this one. [1] https://www.edx.org/course/computation-structures-part-1-digital-circuits-3 https://www.edx.org/course/computation-structures-part-1-dig... [2] https://www.edx.org/course/computation-structures-2-computer-architecture https://www.edx.org/course/computation-structures-2-computer... [3] https://www.edx.org/course/computation-structures-3-computer-organization-2 https://www.edx.org/course/computation-structures-3-computer...

acqq 7y ago
It’s not the logic but the technology that is not widely known: the new fabs cost billions.