Kickoff (I)

Sergio Álvarez - Middleware, Grid Computing, English December 13th, 2006

We are getting ready for our leap towards public releases. What we expect to have just after Christmas is:

  • Our open-source multiplatform middleware, along with the binaries for Windows, Mac, Unix/Linux and others such as PSP or Symbian S60; the documentation in Doxygen and the source code
  • A developer portal in Track for the maintenance and improvement of the middleware
  • A limited version of our framework interface available on the Web.
  • Our corporative web, replacing the dull "visit our blog"

We are working hard in the middleware for this "kickoff". We are migrating it to as many platforms as possible, and working hard on the documentation, in and out of the code. Our goal is to get the grid community enrolled in this project.

Borja pointed out that there was not enough talking about how grid standards are being implemented in our middleware. In fact, the first release will not implement the standards, but we have specified, designed and implemented having OGSA and WSRF in mind. The papers about them have always occupied a main place in our desks. Therefore, we have planned to implement the wrapper - that will give the middleware support for the standards - before February 2007.

Last but not least, in this our first kickoff preview, here are the technological specifications of the middleware:

  • Written in ISO C++
  • Size of the binaries: varies, aprox. 400-500kb in Unix/Linux x86 - aprox. 55kb in PSP 1.5
  • Four resource classes fully available in this release: shared memory, disk storage, process execution, MySQL database (their availability depends on the platform)
  • Installation wizard for Windows, Preference Pane for Mac, configure + make scripts for Unix/Linux

In the next kickoff preview, we’ll talk about the inner design of the middleware. Stay tuned.

“What are all those guys doing down there at 2 a.m.?”

Xavier Fernández - think in grid, English November 17th, 2006

That is what our neighbors wonder about. Here is the answer:

Think in grid was founded in early 2006. We are a group of heterogeneous people located in Barcelona. Young people with fresh ideas and a breakthrough technology that will offer an interesting range of solutions in a few time.

We have developed the first multiplatform multidevice Grid application, which is also as light as 300Kb. We are also in the final steps of developing a framework, which easily allows programmers to develop services that work with it.

To sum up, we have a system that distributes not only processing capacity of the devices involved, but also storage and accessibility to any kind of resource within these devices.

As a friend would do, all the devices in a network work together to provide a service to one or more than one members of this network.

All that stuff is how we do things. Right now we are working in the following fields:

1. Embedded devices: Wherever you are, whatever you need. Your mobile device is no longer a mobile device, but all you want wherever and whenever you want it. Because resources are virtualized, you only need to develop applications once. Take advantage of other devices available to develop application you never imagined. It is not a matter of having all your data, information and files with you, but to be able to access to it, and that is what we do.

2. Streaming: due to the high processing capacity and interoperability that our technology provides, audio and video streaming can be delivered to a huge range of computers and devices, including mobile phones, PSPs and s.o.

3. Massive distributed computing capacity: do you guess which is the less optimized resource in your company? Computers! Only 8 hours working at, let’s say, 10% of what they are capable of? We take advantage of this valuable resource to save you costs while you are processing huge amounts of data you never thought you are capable of. It is not only that it reduces costs, but also allows you to obtain valuable and differentiable information from the data you already have, something that will give you a clear competitive advantage. IBM makes you buy mainframes? Never more!

4. Virtual Organizations and Social Networks.
You should have people and objectives; we deliver the resources and infrastructure needed. Build “all to all” connections, forget about complex networks and extremely pricy solutions.

We are also working in other interesting stuff such as sharing PSP saved games or creating a world community of processing & storage.

Want to know more? Keep visiting us. More and more things are on the way : )

A brief introduction to think in grid

Diego Mariño - think in grid, English October 24th, 2006

Yesterday lots of people tried to translate the content of the post with different web translators. We have tried them and we have checked that their translations are very poor, so let me explain briefly who we we are what and we do.

We are an startup. A great idea, a bunch of geeks, a garage, lots of pizza, even more coffee… you know. We started researching about Grid Computing in a R+D group of the University (UPC) 3 years ago, and we made some projects for telcom and biotechnology firms. Our main finding was checking that the most used solution (Globus) doesn’t suit the needs of most companies.

Our solution is the first Grid application which can run in all the operating systems (Windows, Linux, MacOSX, BSD’s…). It can be used in all kind of devices (computers, PDA’s, mobile phones… even fridges) due to its small size (less than 400 kbytes). Furthermore, it was designed thinking in modularity and escalability, so it’s easy developing stable and light solutions with it.

We are also developing a framework for making  building of grid applications simply. This framework will be able to run complex services transparently, deciding in each moment which resources of the grid are going to be used in each moment, thanks to a layer of intelligent agents. Any kind of users will take profit of the framework. Developers will be able to develop services with our API, independently of the infrastructure it will be used in. System administrators will be able to declare devices and their resources in the grid, implementing security policies. Least  but no last, final users, will be able to run services and monitor the grid, using the API or the framework web interface.

If you want more information do not hesitate contacting dmarino-#-thinkingrid.com.