GPU, a Global Processing Unit / News: Recent posts

The Global Processing Unit Team is investigating how to connect the GPU client to the Gridcoin network, so that people running it are rewarded with Gridcoins.

Unlike in Golem (an Ethereum token), the submission of tasks in GPUs will stay free.
A possible way to connect GPU to the Gridcoin network is to create a proxy BOINC project.

If you would like to fund development of this new GPU feature, you can donate Gridcoins at this address:
SEtxQ4SePHSP2xSfjy4MATvCWGoFMCcahn
At time of writing we already collected 12720 Gridcoins.... read more

We launched the funding campaign for the Global Processing Unit project on Indiegogo for 500'000 dollars.

Please help us to create an Internet Supercomputer which offers computational resources for free to anyone including yourself!

The funding campaign is located at at http://www.indiegogo.com/projects/a-global-processing-unit/x/4524548
Thank you!

I would like to start a big funding campaign on Kickstarter for the GPU project! Goal could be to raise 1 million dollars to finish the GPU II version, currently under development. The project's goal is to create the most flexible Internet supercomputer ever built. Think of Seti@home, but to be used for everyone on the planet.

However, the American, Canadian or English citizenship is required to start an account on Kickstarter. If you would like to help in raising fundings, please contact me at tiziano.mengotti AT gmail.com. Help should include creating the account, managing the campaign and adding content to it. Of course, we would help with lot of material, videos and so on.... read more

0.964 has been released without installer in a portable edition. The main feature of this release is the variable daylength added to the simclimate frontend. Additionally, in the Orbit, Simulation and Analysis frontend it is possible to zoom down to the planet and moon level.

Please note that the Pioneer Anomaly recently got solved without need of new physics. The particular geometry of the antenna dish together with the nuclear power source created the additional acceleration towards the Sun.
More details at this link:
http://www.planetary.org/blogs/bruce-betts/3459.html

GPU is currently undergoing its 4th rewrite, codenamed freedom (3th was named solaris which reached version 0.963, 2th polaris, 1th unnamed). The preview package has version 0.5.0. This package is solely intended for developers as there is still no client working, just a bunch of unit tests.

The package is intended to sketch the whole idea of the next generation GPU. Some people might be disappointed that GPU moves away from peer-to-peer to become a client-server architecture, but we believe this is the only drawback. The highlights of the coming release (probably 6 months from now) are:
- client/server architecture should allow scaling of nodes up to several thousands of clients
- client is in Freepascal and can be compiled on Windows, Linux and Mac OSX
- Server is in PHP/mySQL and runs everywhere where LAMP stack runs
- All software used to develop GPU II is Open Source and available freely
- The plugin architecture of GPU I is retained, although the plugin layer has been rewritten (available in preview) and plugins need to be ported as well
- Communication is entirely over HTTP, workunits can be downloaded, and resulting workunits can be uploaded via HTTP back to the server (available on GPU preview package)
- HTTP makes GPU transparent to most proxies, so that clients can run also in companies and not only at home
- GPU II retains features like channels (to run whiteboard, chat and sensor networks), job management, clustering services, all rewritten in light of the experiences
done with GPU I (available on GPU preview package)
- GPU II will stay a multipurpose distributed computing architecture, so that any client and any server will support any computation (unlike the BOINC architecture)
- Each server should be able to support 40-100 clients according to some initial computations, the network should scale with number of servers
- One superserver will manage the whole network of servers... read more

0.960: climate simulation simulates vulcano Eyjafjallajoekull

GPU 0.960 features a simple climate simulation. With the latest update (0.9607) which can be retrieved with the menu "Tools->Autoupdate to latest version" on the main GPU application, the climate simulation
also shows the path of ashes spread by the vulcano Eyjafjallajökull and the CO2 production of the world population.

However, it is difficult to calibrate the parameters of the model. Anyone can try simple tasks like: having a world with stable temperature at 15.5 Celsius, start a nuclear war, move Earth down to Mercury orbit or up to Iupiter, stop earth rotation or revolution, etc.... read more

This is a preview of what is being built right now. Super C.R.A.P is an acronym for Super Cluster Ready At Processing, core of the system is a Rollin KVM switch with 8 outputs (VGA, mouse, keyboard) bought for 218 CHF (=140 €). Additionally, 8 cables with VGA, mouse and keyboard are needed (each one costs about 15 €). An old 10 MBit Ethernet Switch will provide connectivity to the nodes of the cluster.... read more

The first one is from nanobit who fortunately rejoined the project: to use a script engine (e.g the one of Thexa4 released with 0.9256) as basis for distributed stored procedures. Stored procedures are a well-known thing for people working with databases, they are most of the time SQL procedures stored for convenience in what a database system can administer at best: a table with a column for the row number and another for the lines of the script. Using the distributed WAN database firstly introduced by nanobit in 0.943 joined with the stored procedure concept would allow another way of doing distributed computing with one major advance: it is possible for the user to distribute new code in almost real time.... read more

During last month GPU experienced a short popularity blast with at least 20 nodes online all the time and with peaks of 25-26 nodes. Though this blast finished (now we have 10-15 nodes as usual in the last two years), the blast brought new improvements in the way the community can work together to achieve the goal of a flexible Internet Supercomputer.

Thanks to everclear64, GPU is availabe as torrent and can be downloaded with a Torrent client like ABC (available at http://pingpong-abc.sourceforge.net ). The GPU torrent itself can be downloaded at http://www.tek-zone.com/gpu/GPU.torrent .... read more

GPU in its latest version 0.949 has a new plugin written by Alex Fihman called Monotone Boolean Functions (mbf.dll). The plugin attempts to approximate Dedekind's problem by means of a Monte Carlo approximation. This is the first "real-world" GPU plugin that tries to solve a mathematical problem.
More information on this plugin can be found at http://www.fihman.com/mbf/mbf.html

0.949 also contains Thexa4's Script Engine, which is a novel frontend written in Visual Basic. This frontend adds scripting capabilities to the GPU stack language. It handles GPU answers from different nodes as a set with unique elements, stored into particular variables. The frontend has also loop and echo support. For more about the GPU stack language go to http://gpu.sourceforge.net/virtual.php... read more

Eternity II plugin

On autoupdate there is a new GPU plugin which can help people solving the Eternity II puzzle. Eternity II is a puzzle with 256 pieces, which have to be placed on a 16x16 board. Each squared piece has 4 faces with a particular sign on it. Adiacent faces of neighboured pieces need the same sign.

A 2 million $ price will be paid to the first valid solution submitted to the puzzle creators.... read more

The File Distributor component ( hosted at http://www.gpu-grid.net/file_distributor ) grew into a
complete solution to distribute work units and applications, suitable for small-sized projects.
The GPU plugin integrates with it, and a standalone client is available as well. The PHP set is
now used to collect computer information on the cluster. The hope is that this data will help
to understand the current scalability issues. With 0.946, projects with Single Work Unit set (typical
for rendering projects) are supported.... read more

First, the GPU organization thanks Laurent Wattieaux ( http://www.lwm-media.org )
for mantaining an own FTP server where occasional users can upload the Terragen images
computed by the GPU network.

After the file distributor, a new paradigm ideated and implemented by nanobit (René Tegel)
enters into the GPU arena. Technically implemented as two new plugins called pwansrw.dll and gpu2pwandb.dll,
the P2P WAN database uses environment variables, which can be synchronized between 2 nodes or the entire network by using hashes.
The paradigm is powerful and could be used to exchange results of computations between nodes without having any loss of data.
The distributed P2P WAN DB is described in the document pwandb_paper.odt.doc, to be found in the docs folder.... read more

GPU is a client for distributed computing over P2P.

0.938 is released with a set of PHP scripts (based on preliminary work done by Justin Adams), which allow clients to download work units in a simple manner. The GPU client has an improved 'downloader' plugin, which can deal with the new file distributor. The plugin uses exponential waiting to avoid webserver overload. To get an idea, you can visit the project FD at http://spartacus.is-a-geek.net/file_distributor
More information on FD is also here: http://gpu.sourceforge.net/file_distributor.php... read more

After an internal discussion between team members, we decided to release 0.935 with the unmodified (GPL version 2), and to remove the public released versions beginning from 0.910 up to 0.934.

This for two main reasons: one is that Sourceforge.net hosts only projects that are licensed under the Open Source Definition. The project is not big enough to provide its own CVS and web space.

The second one is that GPL cannot be modified without changing its name. So we should have chosen a name like "No military Use Public License".... read more

What started like a little taunt suddenly got another dimension. The GPU project has modified the GPL license a little by adding Asimov's first law of robotics.

Meanwhile, we have been written be members of the Free Software Foundation, asking us to reconsider the change or at least not violate their copyright by removing the preamble and altering the name. We are aware modifying the GPL is not allowed by the GPL license itself, but did it without bad intentions. We go consider what is appropiate. After all, we're not after a legal conflict with the FSF. Give us some time for internal debate, we'll keep you informed.

The statistics engine was slightly improved (in the latest fix on autoupdate 0.9331) to keep record of all nodes which ever connected to the network.
We are now able to compare our community with the bigger ones in the distributed computing arena. Well, after running the statistics engine for some
months, we collected one year of uptime (which is only uptime, because the network is still idle most of the time). We indexed about 80 GB
of internet data. According to nanobit, the amount of crawled data should be 3/4 times the size of the indexed data. And fortuntately, filters are included by default, so we avoid to crawl quite a lot of crap :-#
Still some nodes can appear twice in the statistics, but the bug will automatically disappear as soon as everyone runs a version equal or higher 0.933.... read more

The monthly release schema went broken due to time constraints and stress at work. Anyway with 0.930 we have a stable release which can run for entire months. The latest changes are on CVS and on autoupdate as usual. To get the changes described below, install 0.930 and then choose from the menu 'Tools' the item 'Autoupdate to
latest version'.

Spanish translation

Alberto Pérez Garcìa-Plaza filed a clean spanish localization file. We thank him a lot :-)
To get Spanish, go to Edit->Settings->Language Options and choose 'Change Language'.... read more

As often in life the release 0.930 did not follow exactly the roadmap with the Orsa plugin. It ended instead with a new screensaver, inspired by the BOINC screensavers. People often online on our network will not be too surprised, to see a 3D network mapper blown to full screen.

At the beginning of January, we released a .zip file containing statistics between November 04 and December 05. Beginning from January 06, we also generate and collect XML statistics, which might be parsed to a later point in time.... read more

While Tiziano works sweat and tears to finish the exiting 'liborsa for pascal and gpu' project, others have put some effort to create the much requested C examples.
A simple C plugin we already had, but we were requested for frontend examples. User edsatori has put a small GUI example using BCB (Borland C builder).
I (Rene) made two examples using minGW/Dev-c++ demonstrating UDP communications with the gpu core. One shows how to register to netmapper or chat messages and how to execute a gpu command and fetch the results. The other is the gpu chat, interactive, but on the console as frontend. Both demonstrate how easy udp works, also in single-threaded applications. As suprise is a port from the famous megahal chat AI as gpu chatbot. Stuff for study for the C programmers who are interested in gpu.

In "Tapping the matrix" [1], a talk given by Carlos Justiniano (founder of Chessbrain), GPU is described as an active project with one big drawback. It is implemented, we quote, in an "obscure Pascal dialect". Ok, this was like a trigger for nanobit:

C/C++ compatibility achieved for plugins

Since the beginning, GPU was intended to load standard dynamic link libraries. We used pointer of chars instead of the very practical Delphi Strings for the communication between core and plugins. Using Bloodshed Dev-C++ [2], an amazing C++ IDE written in Delphi (ehm) and under GPL, nanobit was able to compile a C-DLL. The source code for the C plugin is under /src/dllbuilding/c_example. We had to change a little the TStack definition, so that the parameters were understood by the plugin. This prompted us to recompile all plugins, and recompiling everything was easy step thanks to Samuel Herzogs "DelphiPackageTool". So, we can now easily plug-in any sort of C/C++-code into the GPU system.... read more

GPU is an experimental grid computing project which targets the rise of a big and cheap supercomputer over Internet, run by volunteers.
This is the project status for October 12, 2005.

First, we thank gratefully lwm (Laurent Wattieaux) for adding a new entry host for the GPU network, which solves the problem of the momentaneously broken spartacus node. A big thanks to paulatreides, for fixing some problems in earthsim plugin and reformatting most of the GPU source code with the Jedi formatter. Now the source code looks most of the time nicer and more readable. With time, we will reformat some parts where Jedi did not perform so well. The formatted source code is already on CVS and will be released with 0.925.
Lot of credits to sabybuld (Stefano Godenzi) for buying and planning the water cooling system which will be built in spartacus very soon. Delphifreak (Samuel Herzog) keeps his support to the DelphiPackagerTool, too. pisces computer is now running version 0.9194 for more than 850 hours (35 days) without application crashes.
Red's movie ShipRockFinal was published on the Movies section as well.... read more

GPU is a software for clustering over P2P. This is the project status for 31 August, 2005. Current release is 0.921. Here the areas were there was improvement.

Search Engine

nanobit (Rene Tegel) did a lot to improve the search engine. The new library sqlite3.dll (that replaces sqlite.dll) is able to handle UTF8 characters, so that languages like Chinese are supported. The updated search plugin now does not commit each time a new webpage is crawled but waits for several changes and then commits. As consequence, the journaling of NTFS is much less stressed, and disk usage is reduced a lot. Finally, the webinterface on http://search.dubaron.com that runs mysql is able to work in synergy with the sqlite plugins on the GPU network.... read more

GPU is a cluster software with a P2P based architecture. This is the project status for July, 17, 2005.

Despite summer hole and high temperatures, both the cluster and the development kept running at sustained rhythm.

Network Status

On July 12, <Budtske> connected his Internet Coffe to the GPU system. During that day, 24 nodes formed one unique entity and joined their forces to compute a movie by <Red>. Smoking like an old steam motive, the cluster broke the 30 gigaflops barrier.
http://www.tea.ch/gpu/statistics/july05/gpu_stats_120705_80GHz.html... read more

This is the project status for GPU, a "Gnutella Processing Unit" on May, 22, 2005. GPU is a software able to cluster computers around the internet in a random graph.

Network status

Some problems for the entry points <iap_mohan> and <spartacus> seem solved now. Great thanks to <swiftstream> and <leo> nodes that kept the network running despite all troubles. In order to survive, the GPU network needs more entry points. Read here how to create one: http://gpu.sourceforge.net/advanced.php#publish
Unfortunately, there were no video computations going on these days, too. Additional anonymous ftp servers that could host computed images would be cool... ... read more