Based off some of my research this spring into open chips thusfar everyone is using asic &c software from the 80's
@theruran This is so cool
sweet! The next step is to be able to create an affordable, handmade ultra clean room mod for your garage. Then we're talking big leagues 😉
@rick_777 @theruran You also need to establish safety procedures and waste policies before you get bigger. You don't want to contaminate your backyard with a hundred gallons of toxic TCE in a lab accident.
The dark side of digital revolution - Many standard safety procedures we take for granted didn't exist at all in early days, and almost all the big-name firms, such as Fairchild, National, HP, had accidents back in the days. Silicon Valley remained the most polluted place in US.
@drwho @rick_777 @theruran Yes, the huge 250-gallon underground concrete storage tank full of TCE at MOS/Commodore had a massive leak in 1974, million dollars have been spent to clean it up and today the site is still not entirely safe and continue being monitored by the EPA.
Pretty much a disaster but still interesting history: https://semspub.epa.gov/work/03/2218465.pdf
C128 engineer Bill Herd said he didn't know the tank leaked and parked at the wrong side, and his car was completely contaminated in the cleanup process.
Yeah, that's pretty much it - mechanical logic on the scale of a red blood cell or smaller. With the option of reversible computations, to keep waste heat down.
@deejoe @theruran @niconiconi @rick_777 Technically, we're already using the early stages of nanotech. Samsung is fabbing 8nm chips for its phones. Samsung and TSMC are already fabbing 5nm circuitry. Server class processor cores are already in the 6-7 nm range, which is why we've got processors with 32 and 64 cores on a single die.
On the bio side, we've been patching the DNA of bacteria to synthesize insulin, HGH, interferons, and the vaccine for hepatitis-b. That's also nanotech - modifying DNA to a specific end, and using bacteria to fabricate custom proteins in bulk.
IBM has been using scanning-tunneling microscopes to manipulate individual atoms for twenty-some years now (fuck me...)
Here's a paper from last year about using individual atoms to build geartrains: https://arxiv.org/pdf/1802.01802.pdf
We've gone beyond wishing. We're at the "let's play around to figure out the basics" stage, and have been for a while.
If we're talking purpose-built nano which does only one thing, like push iron atoms around, or build geartrains or something, it's highly unlikely. Sure, there are catastrophic failure modes (there always are) but they're not "eat the world," they're "hey, my knife fell apart."
It's not clear if it'd be possible to build self-replicating nano. I'm not entirely sure it's possible due to the difficulty involved in even fabbing the early generations. It's likely that the toolchain would be something like "Here's a breed of nano that makes a breed of nano that makes synthetic diamond sheets."
We also don't know how nano would have to be controlled. For the forseeable first few generations, it's probably going to use a bicameral architecture - the nano's just the toolkit, but it's controlled by an external (and probably much bigger) mechanism, like a full sized computer. Hit ^c, terminate the control program, and the swarm grinds to a halt because there's no number crunching on board, just a receiver for instructions.
Bionano should be biodegradable. The hacked bacteria we already use to manufacture synthetic hormones in bulk certainly are. Don't refresh their growth medium, stop feeding them, boil them, dump in a handful of drain cleaner crystals, and you kill the entire tank.
I know that we're definitely at the stage of building 3d integrated circuitry: https://en.wikipedia.org/wiki/Three-dimensional_integrated_circuit
Vias are being used in integrated circuitry to not only link layers (each layer being an entire processor core), but as heat pipes for cooling of same.
If we can get process control down to the point where we're working at the 5-10 nm level, we might be able to. I don't know, though.
Start with a Si substrate, grow your layer with either diffusion or direct deposition, etch your extra material away, throw in some implant here and there, and repeat a thousand times.
Here's a video describing the process (in a very simplified manner)
A bunch of technomancers in the fediverse. Keep it fairly clean please. This arcology is for all who wash up upon it's digital shore.