Posts: 62
Threads: 3
Joined: May 2014
Hi, my friends call me Canner IRL so I chose this name online.
A while back while reading some things on the net concerning the theoretical engineering limits of computer chips.
It got me wondering if when we reach the lower limits allowed by physics in building chips if we would be satisfied with the computer power then or still need more and how this could be done. The only solution I could come up with was then make computers get larger and larger since the components had reached the smallest size physically allowed.
Then an idea popped into my head; how large could computers theoretically get and I did some searching on the internet and stumbled across this article that talked about computers the size of Jupiter and needed the power output of an entire star.
There was a few links at the bottom of the article, one took me to OA. Since then I read several dozen articles in the main pages of the part of the site that resembles a theoretical encyclopedia.
If we all get to ask 1 introductory question, this I like to ask if there is an official chat site members use to hang out and talk at?
Posts: 7,299
Threads: 295
Joined: Jan 2013
05-07-2014, 01:02 AM
(This post was last modified: 05-07-2014, 09:24 AM by Rynn.)
Welcome to the forums
![Smile Smile](https://www.orionsarm.com/forum/images/smilies/smile.gif)
IIRC the theoretical limit to computations is something like 1e32cps per kg per 1e-19 seconds. Or in other words if you could collapse 1e19 kgs per second into a black hole and use that for computation you'd generate around 1e50cps. That's a hell of a lot of orders of magnitude above current technology IRL.
In OA there are many computational technologies that allow for greater cps per unit volume/mass. MOTEs and ultimate chips are a common example of physical technology, archai can use mag matter to run plasma processors even better than that. And in the extreme we even have objects such as neutron stars and wormholes used as computers. So there's a long way to develop before bigger is the only option
![Smile Smile](https://www.orionsarm.com/forum/images/smilies/smile.gif)
and in terms of bigness the highest archai utilise tens - hundreds of dyson swarms, Jupiter and moon brains just to think.
In answer to your other question there's no chat that we use but the forums can get pretty busy.
Posts: 15,963
Threads: 730
Joined: Sep 2012
Hi there, Welcome to OA!
To answer your second question first:
As Rynn said, we don't have a chatroom, but the forum can seem like one sometimes if folks get going enough on a given topic. Also, you're welcome to ask as many questions as you like here. We may answer directly or refer you to some resource online or both. But please don't feel that you only get X number of questions and no more.
Re your main question: are you referring to ultimate limits for silicon chips such as we make now or limits to computational hardware of any kind? If the latter, Rynn's answer pretty much covers things, although there are interesting discussions and details to be had if we start getting into things like quantum computing or the like.
If you mean limits of current technology, there are a number of interesting candidate technologies that could replace silicon chips in various ways and keep Moore's Law moving along for a good bit longer.
Hope this helps and once again - Welcome to OA
Todd
Posts: 7,299
Threads: 295
Joined: Jan 2013
05-07-2014, 03:42 AM
(This post was last modified: 05-07-2014, 09:23 AM by Rynn.)
(05-07-2014, 03:09 AM)Drashner1 Wrote: If you mean limits of current technology, there are a number of interesting candidate technologies that could replace silicon chips in various ways and keep Moore's Law moving along for a good bit longer.
Technically Moore's Law might go the way of the dodo (so to speak) but that doesn't mean the end of computer development
![Smile Smile](https://www.orionsarm.com/forum/images/smilies/smile.gif)
just means that it won't be suitable to measure computer development in terms of transistor count, nor expect a regular doubling. Even if development slows because computer architecture is more difficult to improve (doubling transistor count on a 2D surface is a relatively easy improvement) we could see improvements in terms of better software and some developments are potentially radical enough to make a huge leap (e.g. universal memory). So Moore's law will probably die as a trend for computer development, the landscape after might be smooth or might have static periods and spurts.
To be honest I'm partially hoping Moore's law does go away. Abundance is the enemy of efficiency and software bloat seems to be quite common. Operating systems have reached sizes of tens of gigabytes! If software developers cannot rely on ever increasing processor capability to run more complex applications software efficiency would have to be given far higher priority, allowing us to do more with less.
Posts: 62
Threads: 3
Joined: May 2014
I updated my profile with my character bio, I will add more when I think of more to add to the neko cat girl race. My character is more than just a baseline neko, she is augmented with bio tech and acquired improved body parts lengthen her life expectancy, health and increasing her intellect. She is working on increasing her status by .05 S points.
Wants to ask Sun Wu K'ung about the nature of our universe’s level of physical reality.