0034249_094 will be finishing in about 6 hours....

Just picked up 34176_094

entity

34176 was known. It has taken 6 days to gain 2 generations.

34249 is a new one. We now have 15 identified.

Thank you.

Mike

I have an unstuck unit ARP1_0033631_106 running on the 64 bit executable and it has timestep of 24 sec. It will take about 1.5 times a regular 64 bit task, like the first two unstuck units I ran. The last unstuck unit I had was also ran on 64 bit and it took about 2.5 times the normal 64 bit task. It also had a timestep of 24 seconds.

So, there's something odd going on here. Some people are seeing some of these tasks take at least double the time of a normal task run on the same bit executable, while other tasks are taking about 1.5 times the amount of time as a normal task, and yet, for me at least, all the unstuck units have had a 24 second timestep..

Thank you, geophi.

As I indicated above, I would expect 24 sec time steps to require at least 50% more CPU time. However, because of the high system requirements normally, this extra could cause some machines to require even more time.

I am not an expert on computer technicalities, but as a mathematician, I can the possible scenarios where throttling can start or strengthen.

However, someone might be able to come up with a better answer.

Mike

I have:
79.027%; 1d 05:46:24; 1/13/2022 20:03:41; ARP1_0033471_095_0;
For comparison:
85.937%; 11:00:39; 1/14/2022 14:25:50; ARP1_0027543_117_0
Apologies if this is already identified - all the unit numbers have become a blur.
This system has been chugging for quite a while. If it would be helpful, I can also try to identify long runtimes from the results page and post them here.