3 ms·

Everyone's missing his point. Yes you could make a computer but an absolute toy compared to a Jetson. You could reasonably get close to the other technologies i

by djmips 8y ago

Everyone's missing his point. Yes you could make a computer but an absolute toy compared to a Jetson. You could reasonably get close to the other technologies including a fridge compressor given a year of study. Your home built CPU isn't going to touch the Jetson.

lnsru 8y ago
We had a procesor design course at the university. Single participants made 68k complexity designs, groups went to 286 and more complex parts. It was 6-8 hours workload every week for 4 months. A year time is a lot! Manufacturing technology is another issue. Same design will deliver different results on 28 nm and 160 nm nodes.
- llampx 8y ago
  What I wonder about is the 10x factor here - all of these designs have already been invented and are available for perusal. How many people could come up with something like this with nothing but a knowledge of mathematics? I sometimes don't even know if I could invent the wheel if it wasn't already invented.
  - lallysingh 8y ago
    Ref: https://github.com/dugagjin/MIPS https://github.com/dugagjin/MIPS A MIPS implementation in VHDL. Most of the source files are < 50 lines of code. After a little bit of magic, almost any technical field is the application of a few principles. If you saw a rock roll down a hill, you'd probably wish you could roll your own stuff instead of lifting it.
dTal 8y ago
I doubt you'd even get as far as hand-making the brushless motor that powers the compressor. And that's just the start - the fluid dynamics to design an efficient compressor is years of study in itself. It's easy to underestimate the complexity of everyday things. But note that fridges are 100 years old, and today's fridges are much better than the fridges of 100 years ago. It took the collective effort of the human race 100 years to get this far.

3 ms·

Everyone's missing his point. Yes you could make a computer but an absolute toy compared to a Jetson. You could reasonably get close to the other technologies i

by djmips 8y ago

lnsru 8y ago
We had a procesor design course at the university. Single participants made 68k complexity designs, groups went to 286 and more complex parts. It was 6-8 hours workload every week for 4 months. A year time is a lot! Manufacturing technology is another issue. Same design will deliver different results on 28 nm and 160 nm nodes.
- llampx 8y ago
  What I wonder about is the 10x factor here - all of these designs have already been invented and are available for perusal. How many people could come up with something like this with nothing but a knowledge of mathematics? I sometimes don't even know if I could invent the wheel if it wasn't already invented.
  - lallysingh 8y ago
    Ref: https://github.com/dugagjin/MIPS https://github.com/dugagjin/MIPS A MIPS implementation in VHDL. Most of the source files are < 50 lines of code. After a little bit of magic, almost any technical field is the application of a few principles. If you saw a rock roll down a hill, you'd probably wish you could roll your own stuff instead of lifting it.
dTal 8y ago
I doubt you'd even get as far as hand-making the brushless motor that powers the compressor. And that's just the start - the fluid dynamics to design an efficient compressor is years of study in itself. It's easy to underestimate the complexity of everyday things. But note that fridges are 100 years old, and today's fridges are much better than the fridges of 100 years ago. It took the collective effort of the human race 100 years to get this far.