20190329, 19:00  #1 
1976 Toyota Corona years forever!
"Wayne"
Nov 2006
Saskatchewan, Canada
2×3^{2}×11×23 Posts 
New way to multiply
I don't feel like subscribing so if anyone already is on NewScientist feel free to comment:
https://www.newscientist.com/article...berstogether/ 
20190329, 19:32  #2  
6809 > 6502
"""""""""""""""""""
Aug 2003
101×103 Posts
2^{2}·3·11·71 Posts 
Quote:


20190329, 20:29  #3 
Sep 2003
3^{2}×7×41 Posts 
Of little practical use, apparently, except for impossibly large numbers.

20190329, 22:35  #4  
"Robert Gerbicz"
Oct 2005
Hungary
5×17^{2} Posts 
Quote:


20190329, 22:58  #5  
Sep 2016
331 Posts 
Quote:
IOW, given a memory latency/bandwidth curve, (where further away means higher latency + less bandwidth), what's the runtime complexity? Right now, it's not entirely representative of real hardware since that curve is a step function for each of the levels of cache/memory/disk, etc... But let's think (very) big here since the new paper is about impractically big sizes anyway. Given these assumptions:
At larger scales, this won't hold. If you assume that each bit of information requires a minimum (nonzero) amount of mass, then you cannot sustain a 3D structure of fixed density. Because at sufficiently large sizes, that structure will be smaller than its Schwarzschild radius. The densest infinitely sized structure you can have without it collapsing into a black hole I believe is a pole of infinite length. This means that the average latency to get data between any two points of this "large stick of memory" is going to be O(N). This is much worse than O(N^(1/3)). But this is just for latency since we're bounded by the speed of light. Bandwidth seems harder to theorize. Is it possible to transmit an infinite amount of information through a finite amount of space? (a naive extrapolation of simple EM wave addition says it is) p.s. Not being a theoretical physicist, I'm not sure on the assertion that there is a minimum space and mass to storing a bit of information. It may be possible to do some sort of infinite superposition of EM waves or quantum state. Last fiddled with by Mysticial on 20190329 at 23:05 

20190329, 23:16  #6 
"Robert Gerbicz"
Oct 2005
Hungary
5·17^{2} Posts 

20190329, 23:26  #7 
Sep 2016
331 Posts 
I'm obviously not an expert in this area, but I don't think quantum entanglement can allow FTL transmission of information.
From what I understand, any method of transmitting information faster than the speed of light can then be used to travel backwards in time and violate causality. If we allow time travel, then things get complicated enough that I'm not sure I want to think about it.  Come to think of it. I'm not sure the concept of "timecomplexity" in the presence of time travel is that interesting. Because you can take as long as you want to compute the result and then send it back to yourself in the past. Thus taking zero (or negative) time. If you can't "take as long as you want" because you'll run up against the heat death of the universe or something, then you can split it up. Do the computation in chunks and send it back to yourself before the "end" to continue it. Rinse and repeat... Last fiddled with by Mysticial on 20190329 at 23:43 
20190330, 00:08  #8  
"Rashid Naimi"
Oct 2015
Remote to Here/There
2^{6}·31 Posts 
Quantum entanglement can not be used to transmit information FTL. It invoves apparent synchronicity in random aspects of entangled particles.
one thing that involves instant travel in the Micro world is Quantum Tunneling. Here is what TheInfallibleSource has to say about it : Quote:
Link to Video Last fiddled with by a1call on 20190330 at 00:16 

Thread Tools  
Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Multiply Pandigital 2  davar55  Puzzles  3  20130107 20:23 
Multiply Pandigital  davar55  Puzzles  18  20101222 22:07 
Multiply  mgb  Lounge  0  20080728 12:54 
Multiply Perfect Numbers  grandpascorpion  Math  5  20061124 19:47 
Critical bug in Gnu MP FFTmultiply code  ET_  Lounge  3  20040311 16:24 