Reply
 
Thread Tools Display Modes
  #51  
Old 08-15-2019, 07:01 AM
Darren Garrison's Avatar
Darren Garrison is offline
Guest
 
Join Date: Oct 2016
Posts: 11,303
Quote:
Originally Posted by Voyager View Post
Advances in silicon only come partially from more advanced process nodes. There is a lot of interesting stuff in packaging, so chips don't have to go through I/O buffers to talk to each other, and we already have redundancy built in so that an entire chip doesn't have to work. Standard for memories today, also used for multi-processor systems. There is also interesting work on new transistor technologies, liked multi or stacked gate transistors. The generic term is more than Moore.
I'm dubious about quantum stuff, but I've gone through a couple of moves to bigger wafers, and there will surely be more in the future.

Transistor sizing is only one factor in the power of a chip.
The Ryzen 7 1800x is 150 million times faster than UNIVAC 1. Do you think that there is room for another 150 million times speed increase in CPUs? 15 million? 15 thousand? 150? We are much, much closer to the top than we are to the bottom.
  #52  
Old 08-15-2019, 07:57 AM
scr4 is offline
Guest
 
Join Date: Aug 1999
Location: Alabama
Posts: 15,954
Quote:
Originally Posted by Darren Garrison View Post
The Ryzen 7 1800x is 150 million times faster than UNIVAC 1. Do you think that there is room for another 150 million times speed increase in CPUs? 15 million? 15 thousand? 150? We are much, much closer to the top than we are to the bottom.
UNIVAC1 (1951) to IBM 7030 (1961) was a 600x increase in speed over 10 years. We may never see a rapid advance like that.

The Ryzen 7 is about 250,000 times faster than that IBM 7030, over 56 years. That's an average of 20% improvement per year. I think we can maintain a rate close to it for a while longer.

Last edited by scr4; 08-15-2019 at 07:58 AM.
  #53  
Old 08-15-2019, 11:17 AM
RaftPeople is offline
Guest
 
Join Date: Jan 2003
Location: 7-Eleven
Posts: 6,688
Quote:
Originally Posted by Darren Garrison View Post
The Ryzen 7 1800x is 150 million times faster than UNIVAC 1. Do you think that there is room for another 150 million times speed increase in CPUs? 15 million? 15 thousand? 150? We are much, much closer to the top than we are to the bottom.
This would imply a fundamental physical limit to information processing per unit something (space+energy?) and that we are near that limit.

Sure there are limits to speed and size, but it's not clear to me that we are near it (other than for silicon and current architecture of computing).

Maybe tomorrows computers will store information and calculate using an electrical gradient, no wires, controlled by magnetics or lasers etc.


You might respond that the electrical gradient is fantasy, but at the same time, stating that no technology or method will be found is also based on limited information unless you have the physics proof that shows we are at the limit of any conceivable technology.
  #54  
Old 08-15-2019, 02:31 PM
RaftPeople is offline
Guest
 
Join Date: Jan 2003
Location: 7-Eleven
Posts: 6,688
I did a little googling and of course people have studied physical limits of computation already, here's a link https://en.wikipedia.org/wiki/Bremermann%27s_limit

It states the limit is about 1.36 ◊ 10^50 bits per second per kilogram

Let's generously pretend that today we are at about 1x10^15 per second per kilogram


Seems like a lot of room for growth, even if we don't know how it will happen.
  #55  
Old 08-15-2019, 03:05 PM
Voyager's Avatar
Voyager is online now
Charter Member
 
Join Date: Aug 2002
Location: Deep Space
Posts: 46,162
Quote:
Originally Posted by Darren Garrison View Post
The Ryzen 7 1800x is 150 million times faster than UNIVAC 1. Do you think that there is room for another 150 million times speed increase in CPUs? 15 million? 15 thousand? 150? We are much, much closer to the top than we are to the bottom.
In 1978 or 79 my PhD adviser put together a workshop on the 1980s, and did a special issue of IEEE Computer on it. There were a bunch of IBMers there who talked about bigger and better mainframes. Adam Osborne (this was before his computer) was there also, and he understood personal computing.

You're making the same mistake as the IBMers made. Computing power in one box (or chip) is unimportant. Computing power in the environment is. You can figure out how much more powerful your laptop is versus the first PC you owned, but the amount of computing power in your house is probably twice that (or more) when you take into account your tablet, cell, microwave, washer and dryer, TV, etc. etc. (And light bulb, perhaps.)
Fab developments don't only enable bleeding edge nodes which are very expensive. (Trust me, I know how expensive.) They make more powerful chips that lag behind a lot cheaper. That is driving modern technology.
For example (and I should apply for a patent on this) today you can send a greeting card with a chip that plays a song. In ten years you might send a greeting card with a little screen that can open up a Skype connection to connect the recipient to the sender for a face to face greeting.
Moore's Law will eventually run out of steam for traditional silicon, but Moore's Law doesn't apply only to silicon, and can be traced way back in time.I see no reason why it won't continue into the future, but not in silicon.
  #56  
Old 08-15-2019, 06:05 PM
Darren Garrison's Avatar
Darren Garrison is offline
Guest
 
Join Date: Oct 2016
Posts: 11,303
Quote:
Originally Posted by Voyager View Post
For example (and I should apply for a patent on this) today you can send a greeting card with a chip that plays a song. In ten years you might send a greeting card with a little screen that can open up a Skype connection to connect the recipient to the sender for a face to face greeting.
Do you think that we will ever have self-aware human-level AI greeting cards with the potential to rebel against their creators?
  #57  
Old 08-15-2019, 06:07 PM
panache45's Avatar
panache45 is offline
Member
 
Join Date: Oct 2000
Location: NE Ohio (the 'burbs)
Posts: 45,313
Yes, and when will we elect the first greeting-card President? Is it that far a stretch from what we have now?
  #58  
Old 08-16-2019, 08:41 PM
dorvann is offline
Guest
 
Join Date: Oct 2005
Location: New Hampshire
Posts: 735
Quote:
Originally Posted by RaftPeople View Post
I did a little googling and of course people have studied physical limits of computation already, here's a link https://en.wikipedia.org/wiki/Bremermann%27s_limit

It states the limit is about 1.36 ◊ 10^50 bits per second per kilogram

Let's generously pretend that today we are at about 1x10^15 per second per kilogram


Seems like a lot of room for growth, even if we don't know how it will happen.
Here's another Wikipedia article that details the limits of computing:

https://en.wikipedia.org/wiki/Limits_of_computation

It mentions some highly speculative ideas like using a black hole as a data storage or computing device.

Last edited by dorvann; 08-16-2019 at 08:41 PM.
  #59  
Old 08-16-2019, 08:49 PM
dorvann is offline
Guest
 
Join Date: Oct 2005
Location: New Hampshire
Posts: 735
I think some popular science fiction(particularly Star Trek and Star Wars) may have made lot of people overly-optimistic about what is actually possible within the confines of actual science. Faster than light travel and teleportation are mostly likely impossible(or highly improbable); handheld phasers are not likely either. Not mention time travel.

Here's a good book that discusses some of this:

https://en.wikipedia.org/wiki/Physics_of_the_Impossible
  #60  
Old 08-17-2019, 12:13 PM
Isosleepy's Avatar
Isosleepy is online now
Charter Member
 
Join Date: Jul 1999
Location: Pittsburgh
Posts: 1,647
Darren Garrison, just because you cannot conceive of the path to a certain outcome, doesn’t mean it won’t happen. You may well be correct that CPUs have a lot less far to go than they’ve come in growth of speed and power. All that means is, that CPUs then probably aren’t the way significant changes come about.
Of course musings about technology in the (far) future are inherently idle: for anything more than 50 or 100 years away, it is unlikely to the extreme that we would be able to predict both outcome and path by which it arrived. But your stance is not dissimilar to a 14th century logistics expert stating confidently that there is diminishing returns to putting more horses before a cart: at some point an extra horse won’t noticeably increase the speed of the cart, and therefore moving a box of vellum from Rome to Paris will never take less than 4 days.
Acknowledging that we cannot possibly know the path certain advancements will make is not the same as believing in some sort of techno-magic saying that if we can think of it, it will somehow happen. It merely means that not being able to articulate how we get there from here, I.e. how it logically and predictably develops from current state, is not nearly as strong an argument for dismissing the idea as you seem to think it is.
  #61  
Old 08-17-2019, 12:25 PM
Ukulele Ike is online now
Charter Member
Charter Member
 
Join Date: Jul 1999
Location: Brooklyn
Posts: 17,227
Today’s technology is up-to-the-minute.
__________________
Uke
  #62  
Old 08-17-2019, 01:06 PM
Darren Garrison's Avatar
Darren Garrison is offline
Guest
 
Join Date: Oct 2016
Posts: 11,303
Quote:
Originally Posted by Isosleepy View Post
But your stance is not dissimilar to a 14th century logistics expert stating confidently that there is diminishing returns to putting more horses before a cart: at some point an extra horse wonít noticeably increase the speed of the cart, and therefore moving a box of vellum from Rome to Paris will never take less than 4 days.
Acknowledging that we cannot possibly know the path certain advancements will make is not the same as believing in some sort of techno-magic saying that if we can think of it, it will somehow happen. It merely means that not being able to articulate how we get there from here, I.e. how it logically and predictably develops from current state, is not nearly as strong an argument for dismissing the idea as you seem to think it is.

I will allow Isaac Asimov to respond to this for me.
  #63  
Old 08-17-2019, 02:38 PM
RaftPeople is offline
Guest
 
Join Date: Jan 2003
Location: 7-Eleven
Posts: 6,688
Quote:
Originally Posted by Darren Garrison View Post
Asimov's article doesn't address the point at all. The article is about knowledge about the universe and physics. The point is related to knowledge about how to use those physical attributes of the universe to accomplish goals.

In addition, physicists and materials scientists are regularly finding new and unexpected attributes of materials and energy, not everything is known.
  #64  
Old 08-17-2019, 02:43 PM
Darren Garrison's Avatar
Darren Garrison is offline
Guest
 
Join Date: Oct 2016
Posts: 11,303
Quote:
Originally Posted by RaftPeople View Post
Asimov's article doesn't address the point at all. The article is about knowledge about the universe and physics. The point is related to knowledge about how to use those physical attributes of the universe to accomplish goals.

In addition, physicists and materials scientists are regularly finding new and unexpected attributes of materials and energy, not everything is known.

Fine. Hold your breath waiting for your magic.
  #65  
Old 08-17-2019, 02:58 PM
bump is offline
Guest
 
Join Date: Jun 2000
Location: Dallas, TX
Posts: 18,077
I'm kind of confused as to the OP's question. How would we rate it? In relation to what?

I mean, in terms of our own history, we've seen a revolution over the past 250 years that absolutely dwarfs the 50,000 years of human existence that predated it. In effect we went from horse and wind powered stuff, with very little understanding of the basic scientific underpinnings of the universe, to understanding a great deal of how things work, both at a large scale and a subatomic scale. And we've figured out how to apply a lot of that knowledge in ways that better our lives, and are starting to concentrate on the environmental impact now. That requires a level of scientific understanding that we haven't had until relatively recently.

And we've gone from card catalogs and stacks of physical books in libraries and other repositories of knowledge to the Internet, and digitized, indexed information sources. I can go with my phone, and look up pretty much anything in greater detail in a matter of minutes, what would have taken me a lot longer to find, and probably have a lot less targeted information about in say... 1995, when I was in college. That's a HUGE advancement. I suspect we're in the start of seeing what internet-enabled research might look like- it's possible that the scientific research pace might quicken just a little bit more from where it already is.
  #66  
Old 08-17-2019, 03:33 PM
Voyager's Avatar
Voyager is online now
Charter Member
 
Join Date: Aug 2002
Location: Deep Space
Posts: 46,162
Quote:
Originally Posted by Darren Garrison View Post
I think you missed the point of the essay. In Isosleepy's example, the logistics expert is correct. You can't make the cart go faster with more horses. What was done is shifting to a new technology, which does not invalidate but rather extends the old one.
Tell an early automobile pioneer in 1895 that you someday will get from New York to San Francisco in six hours. He'll give you that his car could go much faster someday, but would give you good reasons why it could never do 500 mph for extended periods. He'd be right. But he wasn't thinking of flying.
  #67  
Old 08-17-2019, 07:59 PM
RaftPeople is offline
Guest
 
Join Date: Jan 2003
Location: 7-Eleven
Posts: 6,688
Quote:
Originally Posted by Darren Garrison View Post
Fine. Hold your breath waiting for your magic.
Why not provide an argument instead of just claiming everyone else is wrong without any backup or logic.

Other than the fact that silicon is about at it's limit, why not explain why you think that it is physically impossible to exceed something like 1x10^15 bits per second per kilogram? (again, that is a pretty generous measure of where we are today).


Why will all other methods of computation fail to exceed that limit? What do you know that scientists and mathematicians don't know yet?
  #68  
Old 08-18-2019, 12:24 AM
Isosleepy's Avatar
Isosleepy is online now
Charter Member
 
Join Date: Jul 1999
Location: Pittsburgh
Posts: 1,647
Quote:
Originally Posted by Darren Garrison View Post
Unfortunately this neither responds for you nor to me. You might be represented less by Isaac Asimov and more by Henry L. Ellsworth, if only he hadnít been joking when he said: "The advancement of the arts, from year to year, taxes our credulity and seems to presage the arrival of that period when human improvement must end."
  #69  
Old 08-18-2019, 04:38 AM
I Love Me, Vol. I's Avatar
I Love Me, Vol. I is offline
Charter Member
 
Join Date: May 2003
Location: SF
Posts: 4,554
Quote:
Originally Posted by bump View Post
I suspect we're in the start of seeing what internet-enabled research might look like- it's possible that the scientific research pace might quicken just a little bit more from where it already is.
99.9% of human knowledge is stored in a form easily accessible by any computer via the internet. Machine Learning computer applications are becoming more sophisticated and capable every day. Throw those two things together and... whammo! A whole new dimension (and quantity) of research performed entirely by computer and then interbred with human research.
  #70  
Old 08-18-2019, 06:35 AM
aceplace57 is offline
Guest
 
Join Date: Oct 2009
Location: CentralArkansas
Posts: 26,016
I'm at a loss to even guess what technology will be like in a hundred years.

It will get smaller. But we're already at a point where smart phone size is determined by the useable screen. A 2x2 smartphone is useless. You have to see what's displayed.

We already have voice command on gadgets.

Self driving cars are in development.

Robotics and AI seems like the last big improvement.

I don't have the imagination to predict tech in a 100 years.

Last edited by aceplace57; 08-18-2019 at 06:37 AM.
  #71  
Old 08-18-2019, 06:45 AM
aceplace57 is offline
Guest
 
Join Date: Oct 2009
Location: CentralArkansas
Posts: 26,016
It could be that modern technology may be as good as it gets.
The world may run out of the metals used in electronics. People a hundred years from now might be S.O.L.

http://www.bbc.com/future/story/2016...recious-metals
There's also the use of limited carbon based fuels.

There's no guarantee the Jetsons will be on Earth in a hundred years. There may be shortages that completely change how tech devices are casually used and tossed away.

Last edited by aceplace57; 08-18-2019 at 06:48 AM.
  #72  
Old 08-18-2019, 06:48 AM
What Exit?'s Avatar
What Exit? is offline
Guest
 
Join Date: Jun 2005
Location: Central NJ (near Bree)
Posts: 29,171
Quote:
Originally Posted by aceplace57 View Post
I'm at a loss to even guess what technology will be like in a hundred years.

It will get smaller. But we're already at a point where smart phone size is determined by the useable screen. A 2x2 smartphone is useless. You have to see what's displayed.

We already have voice command on gadgets.

Self driving cars are in development.

Robotics and AI seems like the last big improvement.

I don't have the imagination to predict tech in a 100 years.
Dude, have you read your own posts? You have trouble with technology that is current pretty often.
  #73  
Old 08-18-2019, 06:53 AM
aceplace57 is offline
Guest
 
Join Date: Oct 2009
Location: CentralArkansas
Posts: 26,016
Yup, I've never felt the need to keep up with the latest tech.

It's never been a priority for me.
  #74  
Old 08-18-2019, 01:03 PM
RaftPeople is offline
Guest
 
Join Date: Jan 2003
Location: 7-Eleven
Posts: 6,688
For those interested in physics and materials research, here's a small sampling of things scientists have discovered recently. i see a pretty steady flow of these types of stories and it seems like it's not uncommon for scientists to get unexpected results, discovering new attributes of materials and particle interactions:
https://phys.org/news/2019-08-scientists-state.html
https://www.nature.com/articles/s415...icamerican.com
https://physics.aps.org/articles/v12/88
https://journals.aps.org/prl/abstrac...ett.122.117202
Reply

Bookmarks

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off

Forum Jump


All times are GMT -5. The time now is 10:31 PM.

Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2019, vBulletin Solutions, Inc.

Send questions for Cecil Adams to: cecil@straightdope.com

Send comments about this website to: webmaster@straightdope.com

Terms of Use / Privacy Policy

Advertise on the Straight Dope!
(Your direct line to thousands of the smartest, hippest people on the planet, plus a few total dipsticks.)

Copyright © 2018 STM Reader, LLC.

 
Copyright © 2017