yoyo@home is one of the whitelisted Gridcoin projects and brings existing distributed computing projects to the BOINC world using the BOINC Wrapper. Yoyo is currently running the following sub-projects:
- Siever for the CRU project
- Perfect Cuboid
- Elliptic Curve Factorization (ECM)
- Simulator 5 for evolution@home
- OGR client for distributed.net
With the various tools being used for these sub-projects, I was wondering whether the point distribution would be equal across them and if an optimisation in output would be possible for yoyo@home. Fortunately, I have a large number of work-units available from my calculated RAC series, which gave me the opportunity to test this.
The first step was to create a homogenous data-set to allow a proper comparison. This meant that I had to remove all work-units produced by ARM processors for OGR. OGR is the only sub-project supporting this processor, which provides substantially less points on average compared to Intel and AMD processors.
After exclusion of the WUs, I compiled the overview below with the various sub-projects, the number of samples per sub-project, the points generated, the total run-time and the average amount of points per hour for each sub-project.
The result is really interesting. Basically ECM, Siever and Perfect Cloud generate +/- 50% more points/hour than evolution@home and OGR. This means there is a real opportunity to optimise the output for yoyo@home.
You can change the projects in the preference section on the website selecting the following:
Once you have saved the preferences, you will only receive work-units from these 3 sub-projects and gradually your output and RAC will increase.
Enjoy the extra output and keep on crunching!