Author High Performance Computing  (Read 8808 times)

0 Members and 1 Guest are viewing this topic.

Re: High Performance Computing
« Reply #90 on: December 29, 2016, 09:26:06 PM »
If any of you "Geeks" have Netflix Steaming, ya gotta watch this.  Ya just GOTTA!  It is AWESOME!!!
 ;) ;D

Lo And Behold: Reveries Of The Connected World
https://www.netflix.com/watch/80097363?trackId=13641907&tctx=1%2C0%2Cc3e7ab1f-3788-4f4d-b95b-49d76143d1d5-118166871
I do and I will ma'am. Thanks!  :D

Re: High Performance Computing
« Reply #91 on: December 30, 2016, 07:51:57 AM »
HP is sporting a photonic super computer now that uses a unique form of memory transfer. Interesting.

https://fossbytes.com/hp-supercomputer-8000-times-faster/

http://www.extremetech.com/extreme/196003-hp-reveals-more-details-about-the-machine-linux-os-coming-2015-prototype-in-2016


Oh, and Big Blue is opening up Quantum Computing to the public.  :)

http://m.ibm.com/http/www-03.ibm.com/press/us/en/pressrelease/49661.wss

Say Techno, Im curious as to if you have any thoughts about the self driving cars computer technology?  I wonder about things like, what would happen to all the cars on the roads, running on computers, if we had a power outage or a cyber attack of some sort. I wonder if thought is going into ideas for a power grid/backup web of some sort, that would come into play in the event of "utilities failures".

I have the same concerns for a "robotic workforce".

Re: High Performance Computing
« Reply #92 on: December 30, 2016, 08:21:26 AM »
I'm sure it will be just fine.



Re: High Performance Computing
« Reply #93 on: December 30, 2016, 08:23:56 AM »
I'm sure it will be just fine.



Alright.  :)

Re: High Performance Computing
« Reply #94 on: December 30, 2016, 07:31:35 PM »
I'm sure it will be just fine.


OMG!   That's my mechanic!!! :o

Re: High Performance Computing
« Reply #95 on: December 30, 2016, 08:38:21 PM »
Cool movie. ;)

Official Hacker's Game full Movie 2015.


Re: High Performance Computing
« Reply #96 on: January 12, 2017, 07:16:42 PM »
Say Techno, Im curious as to if you have any thoughts about the self driving cars computer technology?  I wonder about things like, what would happen to all the cars on the roads, running on computers, if we had a power outage or a cyber attack of some sort. I wonder if thought is going into ideas for a power grid/backup web of some sort, that would come into play in the event of "utilities failures".

I have the same concerns for a "robotic workforce".
You would think most of them have double and triple redundancies built in, but I haven't checked the code-base on the remote driving cars. There was an episode of Cyberwar that shows just how easy it is to hack into the regular cars right now. In fact, it was downright simple.

There is so much that could go wrong with self driving cars, that it will take years before they have all the contingencies worked out. It's my belief they are pushing these to market way too quickly. Of course they have the money to pay for lawsuits. (risk / cost-benefit analysis has been done I'm sure,) so from a business stand point, the quicker they can bring them to market and start making a profit, the better for the company.  ::)
The liability for the company is extremely high, so you can bet everything that happens in the cars - everything - will be logged and piped back to a main server - including user interaction (on vehicles with passengers , or a real driver.) OBD3 does this already to a certain extent. (On Board Diagnostic) There is also a plugin module from one of the insurance companies that offers a 'Good Driver Discount' with a pluggin module that tracks everything.

It's inevitable that we will be seeing more and more of the self driving cars, but on the plus side...

Re: High Performance Computing
« Reply #97 on: January 12, 2017, 08:06:10 PM »
Cool movie. ;)

Official Hacker's Game full Movie 2015.

Thanks Starr.    :)

Re: High Performance Computing
« Reply #98 on: January 12, 2017, 08:09:43 PM »
Thank Starr.    :)

My pleasure, Techno. :-*

Re: High Performance Computing
« Reply #99 on: January 12, 2017, 08:19:49 PM »
A new study ... suggests it is possible to bind light to a single electron and create a new form of light. The new light form's properties would lead to circuits that work with packages of photons instead of electrons and allow further research of the quantum physical phenomena governing particles smaller than atoms, on a visible scale.

https://watchers.news/2016/08/06/new-form-of-light-created/


Quantum Light Harvesting Hints at Entirely New Form of Computing.

https://www.technologyreview.com/s/522016/quantum-light-harvesting-hints-at-entirely-new-form-of-computing/


Air Force Tests IBM’s Brain-Inspired Chip...

https://www.technologyreview.com/s/603335/air-force-tests-ibms-brain-inspired-chip-as-an-aerial-tank-spotter/

https://www.fastcompany.com/3032872/fast-feed/ibms-3-billion-investment-in-synthetic-brains-and-quantum-computing

Re: High Performance Computing
« Reply #100 on: January 13, 2017, 03:12:01 AM »
I just had a computer setup with Ubuntu <sp>. This is my first Linux system. Any words of advice?

Re: High Performance Computing
« Reply #101 on: January 13, 2017, 07:52:34 AM »
I just had a computer setup with Ubuntu <sp>. This is my first Linux system. Any words of advice?

I hope to hear how you like it.

Re: High Performance Computing
« Reply #102 on: January 13, 2017, 07:57:15 AM »
This is my first Linux system. Any words of advice?


Don't install Internet Explorer.

Re: High Performance Computing
« Reply #103 on: January 13, 2017, 08:07:37 AM »

Don't install Internet Explorer.

Never ever...on any system.  :o

Re: High Performance Computing
« Reply #104 on: January 13, 2017, 09:40:08 AM »
I just had a computer setup with Ubuntu <sp>. This is my first Linux system. Any words of advice?

What version are you running ?  14.04 LTS, 15.04, 16.10    ...?

With all versions...
1.  Turn on the firewall.
sudo gufw  <enter>    Default settings are good to start.
2. Develop  an alias list and put it in /home/<username>/.bash_aliases     for commonly used commands.
Make sure your .bashrc file in /home/<username>/   has the lines that look like this:

if [ -f ~/.bash_aliases ]; then
    . ~/.bash_aliases
fi

and not like this:

#if [ -f ~/.bash_aliases ]; then
#    . ~/.bash_aliases
#fi


Those HashTags [sic] are actually code comment delineators and the code parser will skip anything on that line of the file. You want that conditional statement uncommented so it redirects to the .bash_aliases file.
If they contain the '#', backspace the # out and save the file.

http://www.linuxnix.com/linux-alias-command-explained-with-examples/

This one might help later on when you understand the terminal better.
http://tldp.org/LDP/abs/html/sample-bashrc.html

Here's some of the more common ones in my list:
===============================================================

alias update='sudo apt-get update'
alias upgrade='sudo apt-get upgrade'
alias upgr8='sudo apt-get update && sudo apt-get upgrade'

alias fetch='sudo apt-get install'

alias clean='sudo apt-get clean'
alias autoclean='sudo apt-get autoclean'

alias autoremove='sudo apt-get autoremove'

alias purger='sudo apt-get --purge remove
alias autopurge='sudo apt-get --purge remove && sudo apt-get autoremove'

alias l='ls -CF'
alias la='ls -A'
alias ll='ls -lh'
alias logout='sudo xfce4-session-logout --logout'
alias ls='ls --color=auto''
alias reboot='sudo reboot'
alias repo='sudo add-apt-repository'
==============================================================

Try to avoid single letter aliases as they could cause problems when typing.

In the *nixes, the terminal is your friend. Fear it not. Actually, once you get used to it, it's much quicker than the GUI programs.

Install Synaptic package manager. It's much better than the software center. (You have your alias list installed - right ? If so...)  Ctrl+Alt+t (launch the terminal) then type:  fetch synaptic <enter>    you will be prompted for your Password. Type it in, then hit enter/return.

3. speed up the system  ->  create a file named:   .gtkrc-2.0      and place it in:   /home/<username>/
Put this in the file:
=============================================================

# Custom Styles

#  ----------- My Additions ----------
gtk-menu-popup-delay = 0
gtk-menu-popdown-delay = 0
gtk-menu-bar-popup-delay = 0
gtk-enable-animations = 0
gtk-timeout-expand = 0
gtk-timeout-initial = 0
gtk-timeout-repeat = 0

=============================================================
and save the file.  Logout or shutdown, then fire back up.

Download and install the restricted audio / video codecs. This will be different between the versions of Ubuntu.
duckduckgo or ixquick it. (Fuck Google)

Do a search for: '?? things to do after installing ubuntu <your version # here>'  Use common sense when following the results of the pages that the query will bring up.

For your first *nix distro, enjoy it, play with it, and have fun. Download and install the programs you want - there are 100's of thousands of them, then make a backup.  Re-Do Bkp is an excellent program, or you could use the built in utility.

Your distro is highly customizable to almost anything you can think of. Just search it out online.
some other search phases to check:
Linux command line tricks
cowsay
fortune
free ebooks BASH
the linux bible
optimizing ubuntu
customizing ubuntu
terminal tricks
advanced terminal scripting
advanced terminal commands
list ubuntu desktop environments
skinning Ubuntu

Prettify and add functionality to your system @  https://www.gnome-look.org/browse/ord/latest/

;)


Re: High Performance Computing
« Reply #106 on: January 16, 2017, 06:02:05 AM »
Starts Today.
Free 20-hour webinar series includes parallel programming, performance optimization, & remote access to advanced servers:

Intel partner Colfax Research is offering a free 20-hour hands-on in-depth training on parallel programming and performance optimization in computational applications on Intel architecture. The first run in 2017 begins January 16, 2017. Broadcasts start at 17:00 UTC (9:00 am in San Francisco, 12:00 noon in New York, 5:00 pm in London, 8:00 pm in Moscow, 10:30 pm in New Delhi, 2:00 am in Tokyo).

https://goparallel.sourceforge.net/colfax-hands-webinar-series-deep-dive-performance-optimization/

System requirements if you're interested in this:
==============================
To join the webinar:
    Windows Vista – Windows 10, Mac OS X 10.6 (Snow Leopard) – Mac OS X 10.11 (El Capitan), Google Chrome OS or Linux, any distribution that supports the Google Chrome Browser (Firefox will not work!)
    Any modern browser for Windows and Mac OS X; Only Google Chrome browser for Linux (Firefox will not work!)
    1 Mbps or better Internet connection
    2 GB or more RAM
    Speakers or a headset
    Recommended: two displays – one for slides, the other for chat and remote access
---------------------------------------------------------------------------------------------------------------------------------------------------------
For remote access:

    Any recent Linux, Mac OS X, or Windows (additional software required for Windows)
    200 Kbps or better Internet connection
    Allowed outgoing SSH traffic (port 22)

Re: High Performance Computing
« Reply #107 on: January 25, 2017, 01:16:45 PM »
Okay.  Here's one for you techy gurus.  'splain this to me.   In lay terms, please.  Thank you. :D

WHAT IS THIS? HUGE MYSTERY OBJECT APPEARS ABOVE ISRAEL.  1.25.17.


Re: High Performance Computing
« Reply #108 on: January 25, 2017, 01:22:45 PM »
Okay.  Here's one for you techy gurus.  'splain this to me.   In lay terms, please.  Thank you. :D

WHAT IS THIS? HUGE MYSTERY OBJECT APPEARS ABOVE ISRAEL.  1.25.17.



Shouldn't this been in the UFO thread?

Re: High Performance Computing
« Reply #109 on: January 25, 2017, 01:33:00 PM »
Shouldn't this been in the UFO thread?
It's not flying, it's falling.  Still a UFO.  I guess.  I don't know.  I'm not a techy guru!!! :o

Re: High Performance Computing
« Reply #110 on: April 05, 2017, 04:16:37 PM »

Re: High Performance Computing
« Reply #111 on: April 05, 2017, 04:50:58 PM »
Arm enters horse in HPC race.
http://www.theregister.co.uk/2017/01/18/arm_met_office_hpc/


Haven't seen the name Cray in awhile. Do they still make supercomputers? Are there still supercomputers?  ???

Re: High Performance Computing
« Reply #112 on: April 05, 2017, 05:05:55 PM »
Haven't seen the name Cray in awhile. Do they still make supercomputers? Are there still supercomputers?  ???

There's one in your smartphone, according to George Noory.

Re: High Performance Computing
« Reply #113 on: April 11, 2017, 06:49:27 AM »
Haven't seen the name Cray in awhile. Do they still make supercomputers? Are there still supercomputers?  ???
They just announced the release of a new model in November.  ;)
https://www.hpcwire.com/off-the-wire/cray-launches-new-xc50-supercomputer/


Re: High Performance Computing
« Reply #115 on: April 22, 2017, 01:13:50 AM »
Thanks for the article, Doc. :)
Very interesting.

Holding to the CPU front, here's an article on the chips being used in the ' OSIRIS-REx ' Satellite .
The site itself is a gold mine of information on CPU's and computing components. It's worth checking out if you're into that sort of thing.
(Look to the ' Catagories ' menu on the left hand side. )

 


OSIRIS-REx: Bringing Back Some of the asteroid Bennu.   http://www.cpushack.com/category/processor-news/ 

"  The Apollo Group  carbonaceous asteroid Bennu is a potential Earth impactor...          ...NASA decided to just go fetch a bit of Bennu.  Thus is the mission of OSIRIS-REx which was launched a few days ago (Sept 8, 2016) aboard an Atlas V 441 as an $850 Million New Frontiers mission.  "



On a different note;  Singularity - Running the Docker alternative utilizing MPI in an HPC environment.
https://www.nextplatform.com/2017/04/10/singularity-containers-hpc-reproducibility-mobility/

Vector processing goes multi-core.
[http://www.nec.com/en/press/201511/global_20151130_01.html      (2015)

What's a vector Super-Computer, you ask ?
https://www.researchgate.net/publication/300588349_The_Brand-New_Vector_Supercomputer_SX-ACE
(This is their 2nd generation product [article from 2014] )

A good write-up from 2008 - on their 1st gen system. (a more in depth explanation)
http://www.nidokidos.org/threads/20389-NEC-Vector-Supercomputer-SX-9-Worlds-Fastest-Supecomputer

Quantum Neuromorphic Deep Learning.  ( That's a mouthful  ;)  )
https://www.nextplatform.com/2017/03/29/neuromorphic-quantum-supercomputing-mesh-deep-learning/

What's Titan ? In 2012, she was the worlds fastest supercomputer - Manufactured by Cray.
http://www.anandtech.com/show/6421/inside-the-titan-supercomputer-299k-amd-x86-cores-and-186k-nvidia-gpu-cores

A little background on Seymour Cray.
http://www.ipwatchdog.com/2015/12/05/seymour-cray-father-of-supercomputers/id=63515/

Re: High Performance Computing
« Reply #116 on: April 22, 2017, 01:51:35 AM »
Thanks for the article, Doc. :)
Very interesting.

Thanks but I'm still not sure what they mean by tensor processing. Any ideas?  ???

Re: High Performance Computing
« Reply #117 on: April 22, 2017, 02:44:29 AM »
Thanks but I'm still not sure what they mean by tensor processing. Any ideas?  ???
If you have a discreet graphics card it is referred to as a GPU (Graphics Processing Unit) and lives on the PCI-e bus of the motherboard. NVidia Tesla's and Intel XEON PHI's are considered co-processors. Basically, you slam a shitload of CPU cores on a gaming card form factor and attach it to the fastest bus on the motherboard. (The PCI Express slot.) It's a way to increase your computing power, but, they generate heat. Think of heat as inefficiency - wasted computing power.
Co-processors can do basically anything the CPU can.
Now, what Google did, was to take ASICs, (Application Specific Integrated Circuit) which differ from the standard CPU cores in that  in a full custom ASIC, you have total control over the size of every transistor forming every logic gate, so you can "fine tune" each gate for optimum performance. A full custom ASIC performs electronic operations as fast as is possible,  providing that the circuit design is efficiently architected. They are also extremely power efficient.
The new Tensor Processing Units, are specifically designed to do only tensor calculations, which means they can fit more transistors on the chip that do only one thing, and do it extremely well--achieving higher efficiency than other types of chips.
ASICs are very well suited for specific purposes. In fact, if you are going to set up a Crypto-Currency mining rig, ASICs are the way to go. But I digress...
Tensor calculations are the only thing the TSU does. It won't accelerate your graphics or give you more frames per second. It is a pure compute module. But, if you are using it in the context of what was described in the article, highly complicated tensor calculations (for deep machine learning), the calculations can be accomplished exponentially quicker and at a fraction of the power draw of a GPU, Field Programmable Gate Arrays, or Vision Processing Units.

Did that help, or did I just muddy the waters ?

Try this...       http://www.techradar.com/news/computing-components/processors/google-s-tensor-processing-unit-explained-this-is-what-the-future-of-computing-looks-like-1326915      ;)

Re: High Performance Computing
« Reply #118 on: April 22, 2017, 03:07:20 AM »
Did that help, or did I just muddy the waters ?

A little of both but let me see if I can unpack it and you tell me if I'm on the right track or not. We've kind of reached the limit of how many transistors we can cram into a silicon chip so now we're trying to cram as many co-processors into that small space as well to extend the computational power and/or functionality of the microprocessor?  ???

Well, whenever I read about an order of magnitude jump that's pretty exciting. It seemed like we we're stalled for awhile.

Re: High Performance Computing
« Reply #119 on: April 22, 2017, 05:41:52 AM »
A little of both but let me see if I can unpack it and you tell me if I'm on the right track or not. We've kind of reached the limit of how many transistors we can cram into a silicon chip so now we're trying to cram as many co-processors into that small space as well to extend the computational power and/or functionality of the microprocessor?  ???

Well, whenever I read about an order of magnitude jump that's pretty exciting. It seemed like we we're stalled for awhile.
That's part of it, but they are also using a different approach. Instead of designing silicon for general purpose use, they are designing it to do one thing only, and to do that one thing very well. In specializing the hardware itself, efficiency in speed and power become easier to realize.
TensorFlow is the name of the specialized software that was designed to run on the TPU.




Found an article that explains it better than I did.  ;)
http://www.techradar.com/news/computing-components/processors/google-s-tensor-processing-unit-explained-this-is-what-the-future-of-computing-looks-like-1326915