USA team member
If someone can help us just a little bit more, our third place will be assured.
Usually when the CPU is loaded 100% processing times per task increase. As you noted this can actually lead to reduced output, but it's dependent on the workload and might not be true on every project. You just have to test and see. The estimated time isn't always correct either, look at the reported task for actual run times for a better idea of what's happening.I found something interesting on my laptop running this challenge. I have an i7-9750H in the machine with 6 cores-12 threads. When running 12 tasks simultaneously the estimated time per work unit was ~1:40 for the SCC work. When I reduced the # of allowed simultaneous tasks to 8 instead of 12 the estimated time per work unit is now ~1:00. I suspected that would mean the total throughput is increased and when I went to look at the device history at WCG I certainly see an uptick in the daily results. Seems that playing with settings for each machine might optimize the performance better than just going full bore with it.
Crunch on. Still in 3rd place
|2||SNURK and friends||152,823,820||-164,454,886||543,597||18,022,240||18,149,930||18,065,281|
|9||Czech National Team||84,189,967||-9,717,755||426,730||10,669,333||10,240,144||10,285,966|