Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.os.linux.advocacy > #379449 > unrolled thread

The fault in our computers

Back to article view | Back to comp.os.linux.advocacy

Contents

  The fault in our computers Takuya Saitoh <taka0038@gmail.com> - 2016-11-06 07:51 -0800
    Re: The fault in our computers Takuya Saitoh <taka0038@gmail.com> - 2016-11-06 07:55 -0800
      Re: The fault in our computers Melzzzzz <mel@zzzzz.com> - 2016-11-06 17:02 +0100
    Re: The fault in our computers Melzzzzz <mel@zzzzz.com> - 2016-11-06 17:03 +0100
      Re: The fault in our computers Takuya Saitoh <taka0038@gmail.com> - 2016-11-06 08:19 -0800
        Re: The fault in our computers Melzzzzz <mel@zzzzz.com> - 2016-11-06 17:24 +0100
        Re: The fault in our computers John Gohde <john.h.gohde@gmail.com> - 2016-11-06 14:52 -0800
    Re: The fault in our computers Adlbifhr Mjduhgfks <am@random.us> - 2016-11-06 18:02 +0000
      Re: The fault in our computers Takuya Saitoh <taka0038@gmail.com> - 2016-11-07 00:17 -0800
      Re: The fault in our computers chrisv <chrisv@nospam.invalid> - 2016-11-07 07:40 -0600
        Re: The fault in our computers Chris Ahlstrom <OFeem1987@teleworm.us> - 2016-11-07 09:22 -0500
          Re: The fault in our computers Bob Glennford <rigormortis@uymail.com> - 2016-11-07 11:19 -0500
            Re: The fault in our computers Chris Ahlstrom <OFeem1987@teleworm.us> - 2016-11-07 12:51 -0500
              Re: The fault in our computers Bob Glennford <rigormortis@uymail.com> - 2016-11-07 19:43 -0500
        Re: The fault in our computers fr314159@gmail.com - 2016-11-07 09:27 -0800
          Re: The fault in our computers chrisv <chrisv@nospam.invalid> - 2016-11-08 07:23 -0600
    Re: The fault in our computers Siri Cruise <chine.bleu@yahoo.com> - 2016-11-07 09:48 -0700

#379449 — The fault in our computers

In short, a big problem with modern computers is that they store data and solve problems with it in two entirely different areas: the memory, and the central processing unit (CPU). And all that shuffling back and fourth takes its toll, says Di Ventra. "To make a quick comparison: our own brain expends about 20 watts to perform 10^16 operations per second," he says, while a supercomputer would require 10 million times more power to do the same number of operations. "A big chunk of that power is wasted in the back and forth transfer of information between the CPU and the memory," says Di Ventra.

Di Ventra's memcomputer sprung out of an easy-to-understand thought experiment from the 1970's. What if, like our brains, a computer stored data in the exact same place it crunched the numbers? And better yet, what if the actual process of crunching data was used as memory?

This type of memory-crunching computer (hence: memcomputer) would sidestep the costly data shuffle. Furthermore, mathematicians have actually proven that that 2-for-1 process would also allow memcomputers to solve certain fantastically complex problems in a single step.

To build his memcomputer, Di Ventra and his colleagues had to physically rebuild and reprogram a computer from its most basic components. Rather than classical silicon transistors (the building blocks that combine to build all electronics), at the core of Di Ventra's machine are what he calls memprocessors. Di Ventra's simple computer uses 6 of them.

Here's how they work. A classical transistor's job basically boils down to one thing, either letting energy through, or not, depending one what it's been told to do. A memprocessor does this exact same job, but it also physically changes some of its properties ("such as its [electrical] resistance," says Di Ventra) depending on how much energy is trying to move through. Even when the memprocessor loses power, it stores that change. In this way, while totally functioning as a classical, data-crunching CPU, memprocessors can also be coded to store resistance-laden information at the same time. No more back and fourth.

MORE: http://www.popularmechanics.com/technology/a16308/memcomputer-quantum-computing-alternative/

http://www.sandiegouniontribune.com/news/science/sdut-memcomputing-diventra-traversa-2015jul03-story.html

[toc] | [next] | [standalone]

#379450

Quantum Computers in Microtubules

One way to consider synchronized neurons involves quantum computers in the microtubules of the neurons.  These quantum calculations throughout the brain would be connected by coherence. This type of microtubule quantum computer would have the unique properties of mind already mentioned, as opposed to a conventional computer.

Inside this microtubule helix, with a diameter of 25 nanometers, the local environment is determined by the specific amino acids in the proteins.  For example, the amino acids phenylalanine and tryptophan both have a ring structure creating a specific quantum effect.  In pockets inside the microtubule where these two amino acids predominate, water will be excluded. There is a specific quantum effect that occurs within water excluding space.  It is this London effect that is proposed to be the basis of the quantum computer in the microtubules.

The proposal includes the fact that there are very large amounts – 108 – of these protein molecules in each neuron. Also, the theory has each oscillating at 107 per second. With 100 billion neurons and considering only a thousand dendrite connections per neuron that would give a brain capacity of 1026 operations per second for a quantum brain computer – quite a huge capacity which might be enough for all of a human mind’s details.

microtubules are very significant in the moment-to-moment function of the neurons, responding rapidly with different scaffolding structures.  They seem to behave as a LEGO computer with properties of a computer type language of spatial structures in the neuron.  These neuronal structures, such as axons and dendrites, are built and rebuilt instantly based upon the needs of the neuron.  The factors influencing the 100,000 synapses of one neuron include response to mental events, including attention, planning, volition, conscious and unconscious thought and emotion.

It has been noted that the placebo effect might demonstrate quantum effects.  A suggestion to a patient changes the neuronal system related to pain as measured both by MRI and by report of the patient. This effect is explained by two competing quantum states, that of a state of probing whether there is pain with an expectation that it is less, and the state of the firing of the pain neurons.  Essentially the patient is noticing what his high expectation is, and the placebo effect takes prominence because of the high expectation.

MORE: http://jonlieffmd.com/blog/human-brain/could-the-brain-and-mind-be-a-quantum-computer-quantum-effects-in-brain-and-mind

[toc] | [prev] | [next] | [standalone]

#379452

On Sun, 6 Nov 2016 07:55:27 -0800 (PST)
Takuya Saitoh <taka0038@gmail.com> wrote:

> Quantum Computers in Microtubules
> 
> One way to consider synchronized neurons involves quantum computers
> in the microtubules of the neurons.  These quantum calculations
> throughout the brain would be connected by coherence. This type of
> microtubule quantum computer would have the unique properties of mind
> already mentioned, as opposed to a conventional computer.
> 

That's based on unproven theory....

-- 
press any key to continue or any other to quit

[toc] | [prev] | [next] | [standalone]

#379454

On Sun, 6 Nov 2016 07:51:12 -0800 (PST)
Takuya Saitoh <taka0038@gmail.com> wrote:

> In short, a big problem with modern computers is that they store data
> and solve problems with it in two entirely different areas: the
> memory, and the central processing unit (CPU). And all that shuffling

Problem is that on CPU die can't be as much memory as needed.
It's not the CPU<->memory problem rather how CPU works...

-- 
press any key to continue or any other to quit

[toc] | [prev] | [next] | [standalone]

#379458

On Monday, November 7, 2016 at 1:03:44 AM UTC+9, Melzzzzz wrote:
> On Sun, 6 Nov 2016 07:51:12 -0800 (PST)
> Takuya Saitoh <taka0038@gmail.com> wrote:
> 
> > In short, a big problem with modern computers is that they store data
> > and solve problems with it in two entirely different areas: the
> > memory, and the central processing unit (CPU). And all that shuffling
> 
> Problem is that on CPU die can't be as much memory as needed.
> It's not the CPU<->memory problem rather how CPU works...

Yeah, yo got to forget all about the present hardware architectures, be it x86 or Arm, 32 or 64 bits.  Just build completely new computing unit from scratch and write completely new OS for it from scratch.  No more hobbyware, DOS, GNU etc....  Perhaps it's easier to use what has been already built in yo head!

https://www.youtube.com/watch?v=AyenRCJ_4Ww

[toc] | [prev] | [next] | [standalone]

#379460

On Sun, 6 Nov 2016 08:19:42 -0800 (PST)
Takuya Saitoh <taka0038@gmail.com> wrote:

> On Monday, November 7, 2016 at 1:03:44 AM UTC+9, Melzzzzz wrote:
> > On Sun, 6 Nov 2016 07:51:12 -0800 (PST)
> > Takuya Saitoh <taka0038@gmail.com> wrote:
> >   
> > > In short, a big problem with modern computers is that they store
> > > data and solve problems with it in two entirely different areas:
> > > the memory, and the central processing unit (CPU). And all that
> > > shuffling  
> > 
> > Problem is that on CPU die can't be as much memory as needed.
> > It's not the CPU<->memory problem rather how CPU works...  
> 
> Yeah, yo got to forget all about the present hardware architectures,
> be it x86 or Arm, 32 or 64 bits.  Just build completely new computing
> unit from scratch and write completely new OS for it from scratch.
> No more hobbyware, DOS, GNU etc....  Perhaps it's easier to use what
> has been already built in yo head!
> 
> https://www.youtube.com/watch?v=AyenRCJ_4Ww
> 

Brain is not digital computer. We don't execute algorithms...
Thinking process is *not algorithm*, rather thinking process *creates
algorithms*.
I saw proof that there is no algorithm to make algorithms in 1987 by my
mathematical logic professor...

-- 
press any key to continue or any other to quit

[toc] | [prev] | [next] | [standalone]

#379541

On Sunday, November 6, 2016 at 11:19:44 AM UTC-5, Takuya Saitoh wrote:
> On Monday, November 7, 2016 at 1:03:44 AM UTC+9, Melzzzzz wrote:
> > On Sun, 6 Nov 2016 07:51:12 -0800 (PST)
> > Takuya Saitoh <taka0038@gmail.com> wrote:
> > 
> > > In short, a big problem with modern computers is that they store data
> > > and solve problems with it in two entirely different areas: the
> > > memory, and the central processing unit (CPU). And all that shuffling
> > 
> > Problem is that on CPU die can't be as much memory as needed.
> > It's not the CPU<->memory problem rather how CPU works...
> 
> Yeah, yo got to forget all about the present hardware architectures, be it x86 or Arm, 32 or 64 bits.  Just build completely new computing unit from scratch and write completely new OS for it from scratch.  No more hobbyware, DOS, GNU etc....  Perhaps it's easier to use what has been already built in yo head!


https://www.youtube.com/watch?v=pWJsFb3JGfA

[toc] | [prev] | [next] | [standalone]

#379479

On Sun, 06 Nov 2016 07:51:12 -0800, Takuya Saitoh wrote:

> In short, a big problem with modern computers is that they store data
> and solve problems with it in two entirely different areas: the memory,
> and the central processing unit (CPU). And all that shuffling back and
> fourth takes its toll
>

That's also known as the "von Neumann Bottleneck."

What is needed is fast memory that can be accessed within one
CPU cycle.  So far, that is only possible with "cache" RAM.
But ALL RAM should be cache RAM, then we could implement the
"Harvard architecture" and have much faster computing.


-- 
GNU/Linux Advocate # 57,635,458,241

Hail Linux!  Hail GNU and the FSF!  Hail Stallman!
--------------------------------------------------------------
"Do I contradict myself?  Very well then, I contradict myself.
I am large; I contain multitudes."

Walt Whitman

[toc] | [prev] | [next] | [standalone]

#379596

On Monday, November 7, 2016 at 3:02:56 AM UTC+9, Adlbifhr Mjduhgfks wrote:
> On Sun, 06 Nov 2016 07:51:12 -0800, Takuya Saitoh wrote:
> 
> > In short, a big problem with modern computers is that they store data
> > and solve problems with it in two entirely different areas: the memory,
> > and the central processing unit (CPU). And all that shuffling back and
> > fourth takes its toll
> >
> 
> That's also known as the "von Neumann Bottleneck."
> 
> What is needed is fast memory that can be accessed within one
> CPU cycle.  So far, that is only possible with "cache" RAM.
> But ALL RAM should be cache RAM, then we could implement the
> "Harvard architecture" and have much faster computing.

Which CPU got the biggest cache on die?  Yeah, Intel !!!  No faggots ARM bullshit...

> 
> -- 
> GNU/Linux Advocate # 57,635,458,241
> 
> Hail Linux!  Hail GNU and the FSF!  Hail Stallman!
> --------------------------------------------------------------
> "Do I contradict myself?  Very well then, I contradict myself.
> I am large; I contain multitudes."
> 
> Walt Whitman

[toc] | [prev] | [next] | [standalone]

#379615

Adlbifhr Mjduhgfks wrote:

>But ALL RAM should be cache RAM, then we could implement the
>"Harvard architecture" and have much faster computing.

While we are dreaming, I want a personal rocket-pack.  One that won't
burn my legs and is cheap to gas-up.

-- 
"RedHat reverted back away from systemd."  -  some piece of shit,
lying shamelessly

[toc] | [prev] | [next] | [standalone]

#379634

chrisv wrote this copyrighted missive and expects royalties:

> Adlbifhr Mjduhgfks wrote:
>
>>But ALL RAM should be cache RAM, then we could implement the
>>"Harvard architecture" and have much faster computing.
>
> While we are dreaming, I want a personal rocket-pack.  One that won't
> burn my legs and is cheap to gas-up.

 LOL

-- 
Q:	Why do mountain climbers rope themselves together?
A:	To prevent the sensible ones from going home.

[toc] | [prev] | [next] | [standalone]

#379650

On Mon, 7 Nov 2016 09:22:50 -0500, Chris Ahlstrom wrote:

> chrisv wrote this copyrighted missive and expects royalties:
> 
>> Adlbifhr Mjduhgfks wrote:
>>
>>>But ALL RAM should be cache RAM, then we could implement the
>>>"Harvard architecture" and have much faster computing.
>>
>> While we are dreaming, I want a personal rocket-pack.  One that won't
>> burn my legs and is cheap to gas-up.
> 
>  LOL

You find that funny?

[toc] | [prev] | [next] | [standalone]

#379665

Bob Glennford wrote this copyrighted missive and expects royalties:

> On Mon, 7 Nov 2016 09:22:50 -0500, Chris Ahlstrom wrote:
>
>> chrisv wrote this copyrighted missive and expects royalties:
>> 
>>> Adlbifhr Mjduhgfks wrote:
>>>
>>>>But ALL RAM should be cache RAM, then we could implement the
>>>>"Harvard architecture" and have much faster computing.
>>>
>>> While we are dreaming, I want a personal rocket-pack.  One that won't
>>> burn my legs and is cheap to gas-up.
>> 
>>  LOL
>
> You find that funny?

Yes.  Not as funny as you, though.

-- 
Q:	How did you get into artificial intelligence?
A:	Seemed logical -- I didn't have any real intelligence.

[toc] | [prev] | [next] | [standalone]

#379692

On Mon, 7 Nov 2016 12:51:02 -0500, Chris Ahlstrom wrote:

> Bob Glennford wrote this copyrighted missive and expects royalties:
> 
>> On Mon, 7 Nov 2016 09:22:50 -0500, Chris Ahlstrom wrote:
>>
>>> chrisv wrote this copyrighted missive and expects royalties:
>>> 
>>>> Adlbifhr Mjduhgfks wrote:
>>>>
>>>>>But ALL RAM should be cache RAM, then we could implement the
>>>>>"Harvard architecture" and have much faster computing.
>>>>
>>>> While we are dreaming, I want a personal rocket-pack.  One that won't
>>>> burn my legs and is cheap to gas-up.
>>> 
>>>  LOL
>>
>> You find that funny?
> 
> Yes.  Not as funny as you, though.

Asshat.

[toc] | [prev] | [next] | [standalone]

#379659

On Monday, November 7, 2016 at 8:40:03 AM UTC-5, chrisv wrote:

> 
> While we are dreaming, I want a personal rocket-pack.  One that won't
> burn my legs and is cheap to gas-up.
> 

Whose dreaming?

Fast and CHEAP memory, that can be accessed in step with the CPU, is
the next technical leap forward.

Notice how I emphasize "cheap."  It would be currently possible
to make ALL RAM as fast as cache RAM but it would also be very
expensive to do so.

[toc] | [prev] | [next] | [standalone]

#379729

fr314159@gmail.com wrote:

>Notice how I emphasize "cheap."  It would be currently possible
>to make ALL RAM as fast as cache RAM but it would also be very
>expensive to do so.

The same thing could be said for personal rocket-packs.

Some things, while possible, will never make sense, because the gains
are not worth the costs.

-- 
'whenever there is touch labor being invested, if it isn't a "zero"
expense, then the 'Linux is free' claim becomes untrue.'  -  lying
asshole "-hh"

[toc] | [prev] | [next] | [standalone]

#379662

In article <8fa783dc-7e00-4688-83a8-f21a908a3156@googlegroups.com>,
 Takuya Saitoh <taka0038@gmail.com> wrote:

> In short, a big problem with modern computers is that they store data and 
> solve problems with it in two entirely different areas: the memory, and the 
> central processing unit (CPU). And all that shuffling back and fourth takes 

You should reactivate a CDC STAR-100, and then the fault would be in our STARs.

-- 
:-<> Siri Seal of Disavowal #000-001. Disavowed. Denied. Deleted.
'I desire mercy, not sacrifice.'
Free the Amos Yee one.

[toc] | [prev] | [standalone]

Back to top | Article view | comp.os.linux.advocacy

csiph-web