Good morning,

This afternoon around 18:00 UTC, we'll begin an extreme stress test of the World Community Grid infrastructure with the help of the OpenPandemics - COVID-19 research team in the Forli Lab. We're grateful for the support and hard work of the Forli Lab team in co-creating and refining this test, and we all look forward to seeing what our entire system can do.

For the purposes of this stress test, we have been given 30,000 batches of work units to run on the GPU version of OpenPandemics - COVID-19. These are real work units that will provide data for the project.

We anticipate this test will take approximately 3 days to run through the 30,000 batches that have been provided. However, the test will end as soon as the 30,000 batches have been processed, whether this processing takes less than 3 days or more than 3 days.

The stress test will involve all parts of the World Community Grid pipeline, from generating batches to post-analysis. This will help us identify bottlenecks, and see where and how we can improve. Below is an outline of the pipeline:

• Researchers identify targets and/or ligands to compute
  
• Researchers create batches of work units to be run by World Community Grid volunteers
  
• World Community Grid downloads work units from the researchers' server
  
• World Community Grid builds work units and load them into BOINC (Berkeley Open Infrastructure for Network Computing)
  
• Volunteer computers and devices download these work units
  
• Volunteer computers process the work units
  
• Volunteer computers upload the files back to World Community Grid servers
  
• World Community Grid validates results
  
• World Community Grid assimilates the results
  
• World Community Grid packages batches into tar files for researchers
  
• World Community Grid uploads the packages to the research server
  
• Researchers re-hydrate the results and place data into their database
  
• Researchers perform analysis on the results
  

Three more important points for those who want to participate in the stress test: 

• After the stress test is complete, we will revert to sending out results at a pace of 2,000 work units every 30 minutes. Depending on the researchers' needs, we may modify this in the future, but for the present our plan is to continue at the 2,000 per 30-minute pace.
  
• GPU work units for OpenPandemics - COVID-19 are designed to run on OpenCL version 1.2 and above. However, there are certain cards that still have issues due to having GPU drivers that aren't 100% compatible with OpenCL 1.2. Most of the issues are with cards that were released before 2016.
  
• Please post any issues or questions in this thread where we can see them more easily, rather than creating new threads that may be harder for us to track.
  

Thanks,
-Uplinger