Hello Everyone,

Very, very soon we'll be officially launching our newest project, and you can start seeing work units today. It's been a long wait and has taken a lot of behind-the-scenes work to get this one ready, so we're even more excited than usual. Before the official launch, we wanted to give you all a heads-up about the technical background and requirements for the project.

Background

The Africa Rainfall Project utilizes the Weather Research & Forecasting (WRF) model for the research application. WRF is a very large, mostly Fortran application and the simulations being run require more resources than are typically used for a World Community Grid project. For this reason, volunteers will not be automatically opted into this project. 

Important Points (Please Read):

• This project is opt-in only. Any volunteers who wish to contribute computing time to this project must log into their account and select the project from the website to receive work. Please review the system requirements page for this project. 
  
• The default setting is one task per machine at a time. In addition, due to the larger than normal system requirements, the project will default to sending one task per machine. If any volunteers wish to increase this value, please see the “How do I limit the number of tasks assigned ... for a specific project?” FAQ.
  
• Work unit size and numbers: Due to the large file sizes of the input and output for this project, the number of tasks sent out per day will be much less than typical projects. For launch, we are starting slow, and while the total number of tasks sent out per day may increase some as we get further in to the project, at this point we anticipate those increases to be minimal.
  

More Info on WRF, Work Units, and Checkpoints

For those not familiar with the WRF application, it is used to simulate weather conditions over a region over a defined time period. The work for this project will be broken into small geographical regions, and in the end each region will be simulated for one calendar year. 

Each individual work unit represents 48 hours calendar time for this simulation. Once a result has been validated for the 48 hours, the output will be used to build the input for the next 48 hours of runtime. The average CPU runtime for an individual workunit in beta testing was about 24 hours. This runtime is also much longer than a typical project. 

Due to limitations in the application, we are only able to take 8 checkpoints during each workunit. Depending on the speed of the host, this could result in long times between checkpoints. For this reason, volunteers who choose to run this project may want to consider using the 'Leave applications in memory while suspended' option in their device profile. 

Please ask any questions in this thread, and as always be sure to read previous questions and answers before posting. Thanks to everyone for your support.

Thanks,
armstrdj

*edited to include third bullet in important points list