The best design was 3.858423% by a Boinc Wrapper user (there is no way to tell who it is)
A big problem with v4.44 was that it was noisy, as we noted when v4.45 was released. So if we wanted to be sure that we had the best designs, we needed to start again with 6tc2, and hopefully the optimizer would take us to a better, or at least different results, and that without the noise giving higher results, another design path might be taken.
The results were different. A lower maximum yield, despite almost 20% more simulations, meant that the the increased accuracy was putting off clients finding their own solutions. It had an amazing start, however, as it was at 1.007752% just 4 days after release.
The search space is in the 10219 range with 73 parameters. The lattice showed a single-stage growth pattern.
The X in the lattice name, replaces the t and also indicates extended time, but also marked it as lattice that required v4.45 at minimum.
This is the 'best' design as recreated in v4.45. The original simulation length as recorded was 57 minutes 05 seconds.
Music is Heartquake in Japan by Eology released under a CC-BY-NC-ND license
Next time, Linac900Ext10td2
- Date of release
- April 1st, 2011
- Date ended
- July 12th, 2011
- Total possible number of permutations
- 10219 (1 with 219 zeros after it)
- Number of Simulations Run
- Total MPTS
- Target Energy Range
- 830.41MeV - 969.64MeV
3.858423 (9589.5 Mpts) [v4.45] <Linac900Ext6Xc2> by boinc (Boinc Wrapper)