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

Groups > alt.folklore.computers > #164885 > unrolled thread

Raspberry Pi 3?

Back to article view | Back to alt.folklore.computers

Contents

  Raspberry Pi 3? "gareth G4SDW GQRP #3339" <no.spam@thank.you.invalid> - 2016-06-10 16:30 +0100
    Re: Raspberry Pi 3? "Osmium" <r124c4u102@comcast.net> - 2016-06-10 11:23 -0500
    Re: Raspberry Pi 3? Michael Black <et472@ncf.ca> - 2016-06-10 12:36 -0400
    Re: Raspberry Pi 3? artie <artie.m@gNOSPAMmail.com> - 2016-06-10 09:41 -0700
      Re: Raspberry Pi 3? "gareth G4SDW GQRP #3339" <no.spam@thank.you.invalid> - 2016-06-10 18:00 +0100
      Re: Raspberry Pi 3? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-10 18:31 +0100
      Re: Raspberry Pi 3? Charlie Gibbs <cgibbs@kltpzyxm.invalid> - 2016-06-10 18:03 +0000
        Re: Raspberry Pi 3? Anne & Lynn Wheeler <lynn@garlic.com> - 2016-06-10 11:28 -0700
        Re: Raspberry Pi 3? "Charles Richmond" <numerist@aquaporin4.com> - 2016-06-11 10:33 -0500
          Re: Raspberry Pi 3? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-11 14:04 -0400
            Re: Raspberry Pi 3? "Charles Richmond" <numerist@aquaporin4.com> - 2016-06-12 09:39 -0500
      Re: Raspberry Pi 3? Bob Martin <bob.martin@excite.com> - 2016-06-11 08:01 +0100
    Re: Raspberry Pi 3? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-10 17:52 +0100
      Re: Raspberry Pi 3? "gareth G4SDW GQRP #3339" <no.spam@thank.you.invalid> - 2016-06-10 18:03 +0100
        Re: Raspberry Pi 3? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-10 19:04 +0100
    Re: Raspberry Pi 3? Bob Eager <news0006@eager.cx> - 2016-06-10 19:18 +0000
      Re: Raspberry Pi 3? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-10 20:56 -0400
        Re: Raspberry Pi 3? Morten Reistad <first@last.name.invalid> - 2016-06-11 11:04 +0200
          Re: Raspberry Pi 3? Bob Eager <news0006@eager.cx> - 2016-06-11 10:47 +0000
          Re: Raspberry Pi 3? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-11 07:51 -0400
        Re: Raspberry Pi 3? Bob Eager <news0006@eager.cx> - 2016-06-11 10:44 +0000
      Re: Raspberry Pi 3? mausg@mail.com - 2016-06-11 10:06 +0000
        Re: Raspberry Pi 3? Morten Reistad <first@last.name.invalid> - 2016-06-11 12:50 +0200
          Re: Raspberry Pi 3? mausg@mail.com - 2016-06-12 08:48 +0000
            Re: Raspberry Pi 3? "Charles Richmond" <numerist@aquaporin4.com> - 2016-06-12 09:37 -0500
              Re: Raspberry Pi 3? Anne & Lynn Wheeler <lynn@garlic.com> - 2016-06-12 09:46 -0700
                Re: Raspberry Pi 3? Anne & Lynn Wheeler <lynn@garlic.com> - 2016-06-12 09:51 -0700
                Re: Raspberry Pi 3? Anne & Lynn Wheeler <lynn@garlic.com> - 2016-06-14 13:07 -0700
              Re: Raspberry Pi 3? Morten Reistad <first@last.name.invalid> - 2016-06-12 18:51 +0200
                Re: Raspberry Pi 3? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-12 17:04 -0400
              Re: Raspberry Pi 3? mausg@mail.com - 2016-06-13 08:04 +0000
                Re: Raspberry Pi 3? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-13 10:00 +0100
        Re: Raspberry Pi 3? JimP <solosam90@gmail.com> - 2016-06-11 12:32 -0500

Page 2 of 2 — ← Prev page 1 [2]

#164906

On Fri, 10 Jun 2016 20:56:46 -0400, J. Clarke wrote:

> In article <ds0i3aFc6elU48@mid.individual.net>, news0006@eager.cx
> says...
>> 
>> On Fri, 10 Jun 2016 16:30:40 +0100, gareth G4SDW GQRP #3339 wrote:
>> 
>> > Just been perusing this and marvelling that for £35  /  $35 you get a
>> > quad core 64 bit ARM architecture and aslo studying the ARM
>> > documentation for that architecture makes me wonder whether the
>> > things we discuss here should get the NG renamed to
>> > alt.pointless.nostalgia? :-)
>> > 
>> > It's flabbergasting that 1Gb or RAM and a 64-bit floating point
>> > processor is available for 1/10th of weekly take-home pay instead of
>> > the 500 years worth of take home pay it would have cost to get such
>> > power in the 1960s.
>> 
>> The comparison I like stems from the old joke about the bandwdith
>> provided by a van loaded with magtapes.
>> 
>> These days it's a van (keeping the playing field level) loaded with
>> microSD cards! Or a 747...
>> 
>> I'm not going to attempt the calculation as it will only lead to more
>> pointless discussion.
> 
> Or consider a containership.  Latency's a bitch though.


http://www.maersk.com/en/hardware/triple-e


-- 
Using UNIX since v6 (1975)...

Use the BIG mirror service in the UK:
 http://www.mirrorservice.org

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

#164905

On 2016-06-10, Bob Eager <news0006@eager.cx> wrote:
> On Fri, 10 Jun 2016 16:30:40 +0100, gareth G4SDW GQRP #3339 wrote:
>
>> Just been perusing this and marvelling that for £35  /  $35 you get a
>> quad core 64 bit ARM architecture and aslo studying the ARM
>> documentation for that architecture makes me wonder whether the things
>> we discuss here should get the NG renamed to alt.pointless.nostalgia? 
>> :-)
>> 
>> It's flabbergasting that 1Gb or RAM and a 64-bit floating point
>> processor is available for 1/10th of weekly take-home pay instead of the
>> 500 years worth of take home pay it would have cost to get such power in
>> the 1960s.
>
> The comparison I like stems from the old joke about the bandwdith 
> provided by a van loaded with magtapes.
>
> These days it's a van (keeping the playing field level) loaded with 
> microSD cards! Or a 747...
>
> I'm not going to attempt the calculation as it will only lead to more 
> pointless discussion.
>
>
>

At what point does it get ridiculous?.. how much data do we realy have, ignoring
the vast realms of b*******, such was government data, porn being posted over
the internet, etc. 


-- 
greymaus.eircom.ie
 .
  .
...

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

#164909

In article <slrnnlnl4l.19p.mausg@Smaus.org>,  <mausg@mail.com> wrote:
>On 2016-06-10, Bob Eager <news0006@eager.cx> wrote:
>> On Fri, 10 Jun 2016 16:30:40 +0100, gareth G4SDW GQRP #3339 wrote:
>>
>>> Just been perusing this and marvelling that for Ł35  /  $35 you get a
>>> quad core 64 bit ARM architecture and aslo studying the ARM
>>> documentation for that architecture makes me wonder whether the things
>>> we discuss here should get the NG renamed to alt.pointless.nostalgia? 
>>> :-)
>>> 
>>> It's flabbergasting that 1Gb or RAM and a 64-bit floating point
>>> processor is available for 1/10th of weekly take-home pay instead of the
>>> 500 years worth of take home pay it would have cost to get such power in
>>> the 1960s.
>>
>> The comparison I like stems from the old joke about the bandwdith 
>> provided by a van loaded with magtapes.
>>
>> These days it's a van (keeping the playing field level) loaded with 
>> microSD cards! Or a 747...
>>
>> I'm not going to attempt the calculation as it will only lead to more 
>> pointless discussion.
>>
>>
>>
>
>At what point does it get ridiculous?.. how much data do we realy have, ignoring
>the vast realms of b*******, such was government data, porn being posted over
>the internet, etc. 

A single, high resolution jpeg image is rarely above 100 megabytes. This would
imply something like 4k x 4k pixels, lots of detail and a low compression factor.

I would guess the normal porn image is way below this. (The message does not
require THAT good an image to convey).

I do have some map scans that go to 150MB, but they are exceptions.

Audio is from 120 k/minute (phone grade) to 40 m/minute(5.1 high resulution).
Broadcase audio is normally around 200-300 kbit/second, or 1.5 - 2 megabyte/minute.

Video is another matter. Plain SD compressed is 4G/hour. HD can go to 15. 1k
to 40, 4k to 100G/hour. That can make a significant dent in storage.

Half of all pictures have been taken the last 2 years. Half of all video
(in terms of resolution) has been taken the last 18 months.

So we are at doubling rates of ~20 months. Which the bandwith growth figures
of the Internet seem to bear out; at ~44% yearly growth.

The copper links to homes and small offices are saturated. But bring in fiber, 
and you can double every 20 months for nearly a century. The national 
backbones (from US to smaller nations like NO/SE) can still increase traffic
by around 100 times without laying much new fiber; and by another 5-10 with
embryoninc technology seeing the light of day right now. (like 500G colours).

-- mrr

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

#164933

On 2016-06-11, Morten Reistad <first@last.name.invalid> wrote:
> In article <slrnnlnl4l.19p.mausg@Smaus.org>,  <mausg@mail.com> wrote:
>>On 2016-06-10, Bob Eager <news0006@eager.cx> wrote:
>>> On Fri, 10 Jun 2016 16:30:40 +0100, gareth G4SDW GQRP #3339 wrote:
>>>
>>>> Just been perusing this and marvelling that for £35  /  $35 you get a
>>>> quad core 64 bit ARM architecture and aslo studying the ARM
>>>> documentation for that architecture makes me wonder whether the things
>>>> we discuss here should get the NG renamed to alt.pointless.nostalgia? 
>>>> :-)
>>>> 
>
> A single, high resolution jpeg image is rarely above 100 megabytes. This would
> imply something like 4k x 4k pixels, lots of detail and a low compression factor.
>
> I would guess the normal porn image is way below this. (The message does not
> require THAT good an image to convey).
>
> I do have some map scans that go to 150MB, but they are exceptions.
>
> Audio is from 120 k/minute (phone grade) to 40 m/minute(5.1 high resulution).
> Broadcase audio is normally around 200-300 kbit/second, or 1.5 - 2 megabyte/minute.
>
> Video is another matter. Plain SD compressed is 4G/hour. HD can go to 15. 1k
> to 40, 4k to 100G/hour. That can make a significant dent in storage.
>
> Half of all pictures have been taken the last 2 years. Half of all video
> (in terms of resolution) has been taken the last 18 months.
>
> So we are at doubling rates of ~20 months. Which the bandwith growth figures
> of the Internet seem to bear out; at ~44% yearly growth.
>
> The copper links to homes and small offices are saturated. But bring in fiber, 
> and you can double every 20 months for nearly a century. The national 
> backbones (from US to smaller nations like NO/SE) can still increase traffic
> by around 100 times without laying much new fiber; and by another 5-10 with
> embryoninc technology seeing the light of day right now. (like 500G colours).
>

I notice (yes!) that webm is getting popular in the porn area. 
On mine, avg about 1mg, of course, that is what I would think 
would be handy.




One is appaled at the aamount of data swirling around. The good side
 is that it will swamp governments attempts to store all data.


-- 
greymaus.eircom.ie
 .
  .
...

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

#164943

<mausg@mail.com> wrote in message news:slrnnlq4v9.19c.mausg@Smaus.org...
> On 2016-06-11, Morten Reistad <first@last.name.invalid> wrote:
>> In article <slrnnlnl4l.19p.mausg@Smaus.org>,  <mausg@mail.com> wrote:
>>>On 2016-06-10, Bob Eager <news0006@eager.cx> wrote:
>>>> On Fri, 10 Jun 2016 16:30:40 +0100, gareth G4SDW GQRP #3339 wrote:
>>>>
>>>>> Just been perusing this and marvelling that for £35  /  $35 you get a
>>>>> quad core 64 bit ARM architecture and aslo studying the ARM
>>>>> documentation for that architecture makes me wonder whether the things
>>>>> we discuss here should get the NG renamed to alt.pointless.nostalgia?
>>>>> :-)
>>>>>
>>
>> A single, high resolution jpeg image is rarely above 100 megabytes. This 
>> would
>> imply something like 4k x 4k pixels, lots of detail and a low compression 
>> factor.
>>
>> I would guess the normal porn image is way below this. (The message does 
>> not
>> require THAT good an image to convey).
>>
>> I do have some map scans that go to 150MB, but they are exceptions.
>>
>> Audio is from 120 k/minute (phone grade) to 40 m/minute(5.1 high 
>> resulution).
>> Broadcase audio is normally around 200-300 kbit/second, or 1.5 - 2 
>> megabyte/minute.
>>
>> Video is another matter. Plain SD compressed is 4G/hour. HD can go to 15. 
>> 1k
>> to 40, 4k to 100G/hour. That can make a significant dent in storage.
>>
>> Half of all pictures have been taken the last 2 years. Half of all video
>> (in terms of resolution) has been taken the last 18 months.
>>
>> So we are at doubling rates of ~20 months. Which the bandwith growth 
>> figures
>> of the Internet seem to bear out; at ~44% yearly growth.
>>
>> The copper links to homes and small offices are saturated. But bring in 
>> fiber,
>> and you can double every 20 months for nearly a century. The national
>> backbones (from US to smaller nations like NO/SE) can still increase 
>> traffic
>> by around 100 times without laying much new fiber; and by another 5-10 
>> with
>> embryoninc technology seeing the light of day right now. (like 500G 
>> colours).
>>
>
> I notice (yes!) that webm is getting popular in the porn area.
> On mine, avg about 1mg, of course, that is what I would think
> would be handy.
>
>
>
>
> One is appaled at the aamount of data swirling around. The good side
> is that it will swamp governments attempts to store all data.
>
>

You must *not* have heard of the new *huge* data center for the NSA that 
opened in Utah...

https://www.wired.com/2012/03/ff_nsadatacenter/


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




-- 

numerist at aquaporin4 dot com

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

#164950

"Charles Richmond" <numerist@aquaporin4.com> writes:
> You must *not* have heard of the new *huge* data center for the NSA
> that opened in Utah...
>
> https://www.wired.com/2012/03/ff_nsadatacenter/
>
>
> https://en.wikipedia.org/wiki/Utah_Data_Center

there is the rapidly growing "success of failure" culture
http://www.govexec.com/excellence/management-matters/2007/04/the-success-of-failure/24107/

some past posts
http://manana.garlic.com/~lynn/submisc.html#success.of.failure

and huge uptic in outsourcing last decade (in just these agencies, 70%
of budget and over half the people)
http://www.investingdaily.com/17693/spies-like-us/

we've made snide remarks that the failures are just excuse for enormous
increases in contracts for appropriations to technology related
for-profit commercial operations (if the amount of spending doesn't fix
the problem, then spend more).

The "spies like us" reference that "private equity" operations ... some
past posts
http://manana.garlic.com/~lynn/submisc.html#private.equity

have heavily moved into doing LBO takeover of gov. contractors ... the
outsourcing and private equity has some possible overlapping motivations
1) gov. agencies can't spend budget on congress, 2) gov. contractors
can't spend money from gov. contracts on congress; however private
equity owners can spend all sorts of money on congress (and it is
ambiguous whether the money actually comes from gov. contractors)
... and there is enormous amount of money to be skimmed off from
gov. contractors (like article references security clearances were
outsourced ... and they were just filling out the paper work and not
actually doing background checks ... making more money available to
their private equity owners)

Note that former president of AMEX after he leaves as head of IBM,
becomes head of large private equity company ... and will do LBO
take-over of company mentioned in the "spies like us" article and
employer of Snowden.
http://manana.garlic.com/~lynn/submisc.html#guerstner

Standard process going back to at least 70s-80s (in financial and gov)
is multi-party operations for sensitive operations (as countermeasure to
insider threats). After the Snowden event, I raised the issue of what
happened to multi-party operations.  A few weeks later there was some
press item that the gov.  was looking at (returning to?) multi-party
operations for sensitive operations. One question about not having
multi-party operation was that (also) a cost cuting move by the
gov. contractor??

-- 
virtualization experience starting Jan1968, online at home since Mar1970

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

#164951

Anne & Lynn Wheeler <lynn@garlic.com> writes:
> http://manana.garlic.com/~lynn/submisc.html#guerstner

finger slip
http://manana.garlic.com/~lynn/submisc.html#gerstner

-- 
virtualization experience starting Jan1968, online at home since Mar1970

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

#165125

Anne & Lynn Wheeler <lynn@garlic.com> writes:
> Standard process going back to at least 70s-80s (in financial and gov)
> is multi-party operations for sensitive operations (as countermeasure to
> insider threats). After the Snowden event, I raised the issue of what
> happened to multi-party operations.  A few weeks later there was some
> press item that the gov.  was looking at (returning to?) multi-party
> operations for sensitive operations. One question about not having
> multi-party operation was that (also) a cost cuting move by the
> gov. contractor??

re:
http://manana.garlic.com/~lynn/2016d.html#68 Raspberry Pi 3?
http://manana.garlic.com/~lynn/2016d.html#71 Raspberry Pi 3?


DoD Expands Concerning Insider Threat Program to Contractors
http://www.pogo.org/blog/2016/06/dod-expands-concerning-insider-threat-program.html

While this change addresses a valid concern that classified
information pertinent to our national security must be protected, the
Project On Government Oversight and other good governance
organizations fear it will be used to target whistleblowers. ITPs
threaten whistleblowers' ability to report waste, fraud, and abuse
because they do not differentiate between genuine threats and those
who are acting in the interest of the public

... snip ... 

aka, embarrassing gov. agencies is considered a worse threat than any
foreign adversaries.

whistleblower posts
http://manana.garlic.com/~lynn/submisc.html#whistleblower

also from article

In a PowerPoint presentation given by the Office of the Director of
National Intelligence (ODNI) at their internal insider-threat webinar,
titled "Simple Steps and Guidance to Secure Classified Networks," ODNI
placed NSA whistleblower Thomas Drake in the same "insider threat"
category as the Fort Hood and Navy Yard shooters, Nidal Hasan and Aaron
Alexis. The same PowerPoint defined insider threats as "any employees
and contractors who damage an entity's reputation...by exposing inside
information." By this definition, there is no difference between
whistleblowers who expose government wrongdoing in hopes of fixing
problems, like Drake, and people wishing to do harm to national
security, like Hasan and Alexis.

... snip ...


aka even when following law requiring reporting to responsible
congressional oversight committee and for that, was charged under the
same statutes used to charge Snowden. Drake's charges were eventually
dropped ... although some of the information eventually shows up in the
"success of failure" series of articles
http://www.govexec.com/excellence/management-matters/2007/04/the-success-of-failure/24107/
"success of failure" posts
http://manana.garlic.com/~lynn/submisc.html#success.of.failure

-- 
virtualization experience starting Jan1968, online at home since Mar1970

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

#164952

In article <njjs2p$j0h$1@dont-email.me>,
Charles Richmond <numerist@aquaporin4.com> wrote:
><mausg@mail.com> wrote in message news:slrnnlq4v9.19c.mausg@Smaus.org...
>> On 2016-06-11, Morten Reistad <first@last.name.invalid> wrote:
>>> In article <slrnnlnl4l.19p.mausg@Smaus.org>,  <mausg@mail.com> wrote:
>>>>On 2016-06-10, Bob Eager <news0006@eager.cx> wrote:
>>>>> On Fri, 10 Jun 2016 16:30:40 +0100, gareth G4SDW GQRP #3339 wrote:
>>>>>
>>>>>> Just been perusing this and marvelling that for Ł35  /  $35 you get a
>>>>>> quad core 64 bit ARM architecture and aslo studying the ARM
>>>>>> documentation for that architecture makes me wonder whether the things
>>>>>> we discuss here should get the NG renamed to alt.pointless.nostalgia?
>>>>>> :-)
>>>>>>
>>>
>>> A single, high resolution jpeg image is rarely above 100 megabytes. This 
>>> would
>>> imply something like 4k x 4k pixels, lots of detail and a low compression 
>>> factor.
>>>
>>> I would guess the normal porn image is way below this. (The message does 
>>> not
>>> require THAT good an image to convey).
>>>
>>> I do have some map scans that go to 150MB, but they are exceptions.
>>>
>>> Audio is from 120 k/minute (phone grade) to 40 m/minute(5.1 high 
>>> resulution).
>>> Broadcase audio is normally around 200-300 kbit/second, or 1.5 - 2 
>>> megabyte/minute.
>>>
>>> Video is another matter. Plain SD compressed is 4G/hour. HD can go to 15. 
>>> 1k
>>> to 40, 4k to 100G/hour. That can make a significant dent in storage.
>>>
>>> Half of all pictures have been taken the last 2 years. Half of all video
>>> (in terms of resolution) has been taken the last 18 months.
>>>
>>> So we are at doubling rates of ~20 months. Which the bandwith growth 
>>> figures
>>> of the Internet seem to bear out; at ~44% yearly growth.
>>>
>>> The copper links to homes and small offices are saturated. But bring in 
>>> fiber,
>>> and you can double every 20 months for nearly a century. The national
>>> backbones (from US to smaller nations like NO/SE) can still increase 
>>> traffic
>>> by around 100 times without laying much new fiber; and by another 5-10 
>>> with
>>> embryoninc technology seeing the light of day right now. (like 500G 
>>> colours).
>>>
>>
>> I notice (yes!) that webm is getting popular in the porn area.
>> On mine, avg about 1mg, of course, that is what I would think
>> would be handy.
>>
>>
>>
>>
>> One is appaled at the aamount of data swirling around. The good side
>> is that it will swamp governments attempts to store all data.
>>
>>
>
>You must *not* have heard of the new *huge* data center for the NSA that 
>opened in Utah...
>
>https://www.wired.com/2012/03/ff_nsadatacenter/
>
>
>https://en.wikipedia.org/wiki/Utah_Data_Center

That is just a little larger than Facebook's current site in Lulea in
Northern Sweden. They say "exabytes or more"; so let us assume 10 exabytes
worth of storage. (Which would push the envelope for what is possible
with the allocated incoming power).

At ~50 terabits/second (a guess to the traffic passing through
the US, with about half as average) you are talking about 3-4 terabytes
per second, 24/7/365. Or a third of a petabyte per day, giving that
exabyte in around a decade.

Yep, it seems it all goes into storage there.

But actually accessing it is another matter. Anything more than 
a few tens of millionths in actual processing for analysis would break
the bank on available power. But that seems like a rational approach
given what mandates the NSA has, at least publicly. Store it all, 
but be selective about what is analysed in depth.

-- mrr 

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

#164953

In article <9hg03d-p35.ln1@sambook.reistad.name>, 
first@last.name.invalid says...
> 
> In article <njjs2p$j0h$1@dont-email.me>,
> Charles Richmond <numerist@aquaporin4.com> wrote:
> ><mausg@mail.com> wrote in message news:slrnnlq4v9.19c.mausg@Smaus.org...
> >> On 2016-06-11, Morten Reistad <first@last.name.invalid> wrote:
> >>> In article <slrnnlnl4l.19p.mausg@Smaus.org>,  <mausg@mail.com> wrote:
> >>>>On 2016-06-10, Bob Eager <news0006@eager.cx> wrote:
> >>>>> On Fri, 10 Jun 2016 16:30:40 +0100, gareth G4SDW GQRP #3339 wrote:
> >>>>>
> >>>>>> Just been perusing this and marvelling that for £35  /  $35 you get a
> >>>>>> quad core 64 bit ARM architecture and aslo studying the ARM
> >>>>>> documentation for that architecture makes me wonder whether the things
> >>>>>> we discuss here should get the NG renamed to alt.pointless.nostalgia?
> >>>>>> :-)
> >>>>>>
> >>>
> >>> A single, high resolution jpeg image is rarely above 100 megabytes. This 
> >>> would
> >>> imply something like 4k x 4k pixels, lots of detail and a low compression 
> >>> factor.
> >>>
> >>> I would guess the normal porn image is way below this. (The message does 
> >>> not
> >>> require THAT good an image to convey).
> >>>
> >>> I do have some map scans that go to 150MB, but they are exceptions.
> >>>
> >>> Audio is from 120 k/minute (phone grade) to 40 m/minute(5.1 high 
> >>> resulution).
> >>> Broadcase audio is normally around 200-300 kbit/second, or 1.5 - 2 
> >>> megabyte/minute.
> >>>
> >>> Video is another matter. Plain SD compressed is 4G/hour. HD can go to 15. 
> >>> 1k
> >>> to 40, 4k to 100G/hour. That can make a significant dent in storage.
> >>>
> >>> Half of all pictures have been taken the last 2 years. Half of all video
> >>> (in terms of resolution) has been taken the last 18 months.
> >>>
> >>> So we are at doubling rates of ~20 months. Which the bandwith growth 
> >>> figures
> >>> of the Internet seem to bear out; at ~44% yearly growth.
> >>>
> >>> The copper links to homes and small offices are saturated. But bring in 
> >>> fiber,
> >>> and you can double every 20 months for nearly a century. The national
> >>> backbones (from US to smaller nations like NO/SE) can still increase 
> >>> traffic
> >>> by around 100 times without laying much new fiber; and by another 5-10 
> >>> with
> >>> embryoninc technology seeing the light of day right now. (like 500G 
> >>> colours).
> >>>
> >>
> >> I notice (yes!) that webm is getting popular in the porn area.
> >> On mine, avg about 1mg, of course, that is what I would think
> >> would be handy.
> >>
> >>
> >>
> >>
> >> One is appaled at the aamount of data swirling around. The good side
> >> is that it will swamp governments attempts to store all data.
> >>
> >>
> >
> >You must *not* have heard of the new *huge* data center for the NSA that 
> >opened in Utah...
> >
> >https://www.wired.com/2012/03/ff_nsadatacenter/
> >
> >
> >https://en.wikipedia.org/wiki/Utah_Data_Center
> 
> That is just a little larger than Facebook's current site in Lulea in
> Northern Sweden. They say "exabytes or more"; so let us assume 10 exabytes
> worth of storage. (Which would push the envelope for what is possible
> with the allocated incoming power).

They're saying 65 MW IIRC.  According to Seagate 10 exabytes of single-
redundant storage will draw about 22.5 MW.
 
> At ~50 terabits/second (a guess to the traffic passing through
> the US, with about half as average) you are talking about 3-4 terabytes
> per second, 24/7/365. Or a third of a petabyte per day, giving that
> exabyte in around a decade.
> 
> Yep, it seems it all goes into storage there.
> 
> But actually accessing it is another matter. Anything more than 
> a few tens of millionths in actual processing for analysis would break
> the bank on available power. But that seems like a rational approach
> given what mandates the NSA has, at least publicly. Store it all, 
> but be selective about what is analysed in depth.
> 
> -- mrr 



---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

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

#164955

On 2016-06-12, Charles Richmond <numerist@aquaporin4.com> wrote:
><mausg@mail.com> wrote in message news:slrnnlq4v9.19c.mausg@Smaus.org...
>> On 2016-06-11, Morten Reistad <first@last.name.invalid> wrote:
>>> In article <slrnnlnl4l.19p.mausg@Smaus.org>,  <mausg@mail.com> wrote:
>> would be handy.
>>
>>
>>
>>
>> One is appaled at the aamount of data swirling around. The good side
>> is that it will swamp governments attempts to store all data.
>>
>>
>
> You must *not* have heard of the new *huge* data center for the NSA that 
> opened in Utah...
>
> https://www.wired.com/2012/03/ff_nsadatacenter/
>
>
> https://en.wikipedia.org/wiki/Utah_Data_Center
>
>
>
>

Someone is doing well out of all that. Probably illegal aliens 
doing the grunt work.


-- 
greymaus.eircom.ie
 .
  .
...

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

#164957

On 13 Jun 2016 08:04:49 GMT
mausg@mail.com wrote:

> On 2016-06-12, Charles Richmond <numerist@aquaporin4.com> wrote:
> ><mausg@mail.com> wrote in message news:slrnnlq4v9.19c.mausg@Smaus.org...
> >> On 2016-06-11, Morten Reistad <first@last.name.invalid> wrote:
> >>> In article <slrnnlnl4l.19p.mausg@Smaus.org>,  <mausg@mail.com> wrote:
> >> would be handy.
> >>
> >>
> >>
> >>
> >> One is appaled at the aamount of data swirling around. The good side
> >> is that it will swamp governments attempts to store all data.
> >>
> >>
> >
> > You must *not* have heard of the new *huge* data center for the NSA
> > that opened in Utah...
> >
> > https://www.wired.com/2012/03/ff_nsadatacenter/
> >
> >
> > https://en.wikipedia.org/wiki/Utah_Data_Center
> >
> >
> >
> >
> 
> Someone is doing well out of all that. Probably illegal aliens 
> doing the grunt work.

	Someone is selling a lot of discs, and associated gubbins -
commission on the deal would be nice, even a very small one.

-- 
Steve O'Hara-Smith                          |   Directable Mirror Arrays
C:>WIN                                      | A better way to focus the sun
The computer obeys and wins.                |    licences available see
You lose and Bill collects.                 |    http://www.sohara.org/

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

#164917

On 11 Jun 2016 10:06:29 GMT, mausg@mail.com wrote:

>On 2016-06-10, Bob Eager <news0006@eager.cx> wrote:
>> On Fri, 10 Jun 2016 16:30:40 +0100, gareth G4SDW GQRP #3339 wrote:
>>
>>> Just been perusing this and marvelling that for £35  /  $35 you get a
>>> quad core 64 bit ARM architecture and aslo studying the ARM
>>> documentation for that architecture makes me wonder whether the things
>>> we discuss here should get the NG renamed to alt.pointless.nostalgia? 
>>> :-)
>>> 
>>> It's flabbergasting that 1Gb or RAM and a 64-bit floating point
>>> processor is available for 1/10th of weekly take-home pay instead of the
>>> 500 years worth of take home pay it would have cost to get such power in
>>> the 1960s.
>>
>> The comparison I like stems from the old joke about the bandwdith 
>> provided by a van loaded with magtapes.
>>
>> These days it's a van (keeping the playing field level) loaded with 
>> microSD cards! Or a 747...
>>
>> I'm not going to attempt the calculation as it will only lead to more 
>> pointless discussion.
>>
>>
>>
>
>At what point does it get ridiculous?.. how much data do we realy have, ignoring
>the vast realms of b*******, such was government data, porn being posted over
>the internet, etc. 

Spam emails make up 75-90 percent of all emails.

-- 
JimP.

[toc] | [prev] | [standalone]

Page 2 of 2 — ← Prev page 1 [2]

Back to top | Article view | alt.folklore.computers

csiph-web