Welcome to the official homepage for the HPVM project. Here you will find the latest news, updates, and developer information for the ever more prevalent API for distributed task programming.
The price of "legacy" systems is dropping at a fast rate. Quad PIII servers can be purchased on Ebay for less than $800. Stacks of old SGI systems (Indy and Indigo) systems are being sold for rock bottom prices. But until one single unified API is created to unite all these systems into one single cluster for use by the common computer user, they will remain useless to most people. This is where HPVM comes in. HPVM exists to bring the cluster concept out of the Universities and research facilities and make them useful for everything from video editing, to database servers.
After working with the current network code, and trying to get all the bugs out, it was decided to scrap it all an start over from scratch. The new code will be way more mult-threaded, and is already showing it'self to be more stable. We are basing it off of the Hawk Network Library which has been found to be robust and easy to port and use.
The sourcecode for HPVM is now online! It is now possible to download the lastest version of the software. CVS is currently unavailable (due to firewall and other issues), this may be resolved in the future. But in the mean time download the source and get going! Read the README file for information on how to get started.
The web page is finally up and running, thanks to the generous people who run Sourceforge. With this page update, we are also announcing the successful porting of the HPVM to the OSX platform. Everything is up and running correctly now in Win32, Linux, and OSX. There is only one big step left before the project is ready for the beta release though, and that is adding support for the Kernel Trees. Currently no platform can run kernels properly due to the fact that a Win32 server will be sending a Win32 kernel to a OSX client. So support for this needs to be added. However, support for a Win32 to Win32 kernel is working. As well as file and data transfer between OSX and Linux, etc.
The current version of HPVM is somewhat limited. It is designed to be used in a client/server situation where the client receives instructions from the server. However as the project matures this may change to become more of a peer-to-peer style network that will allow each node to spawn tasks, as well as allow each node to act as a memory server, allowing the total amount of memory usable to each node equal the total of all the available memory on all the nodes combined (granted the latency will be high though).