Recent content by mstenholm

When I fired my 32 treads/32 GB memory PC up I was pleased to noticed that all the new jobs I got was the easy going low memory consuming jobs. Then an hour later I started my 24 thread/32 GB PC and that machine downloads 51 jobs in one go, not like the other that get 3 every 30 seconds. Well...

Rosetta is best (most points, least amount of wasted CPU-time) done 24/24. If you as me turn the PC of in peak hours to avoid paying premium kWh price you risk to lose up to 2 hour on any RosettaVS_SAVE_ALL_OUT_NOJRAN_CHIP since it will restart with 0 seconds completed if you haven't reached the...

And later the same day that number jumped to 2,400,000 and with just 170,000 completed in the past 24 hours we have plenty. Apparently some people beside me found out the now we have three different jobs - 1,6 and 2,6 GB memory use or the new and improved 8a_hal_c_. Take your poision ;). The new...

Download and run Zentimmings, that will give us a starting point for your RAM.

Well if you have zero validated results then the points for all is zero. ;) The validator went back online +12 hours ago and the backlog is cleared so expect a normal but larger update around 8-9 PM UTC.

It's time to seriously reduce my production. I don't need the heat, the current jobs especially RosettaVS_SAVE_ALL_OUT_NOJRAN_CHIP_ have so few if any checkpoint that a restart is from zero and the 2 or 1.5 GB memory size and points all over is not helping either. So when the wind is blowing + 8...

@phill I suspect that this table and the one before is for WCG. Remember that +200K is a good day for our Rosetta team and that the Rosetta server has been down since the morning and out of Free-DC reach.

It seems like the validator is back on Received 6 Apr 2024, 5:24:13 UTC Server state Over Outcome Success

You missed my link? Yes it seems to be the case. They were told 30 hours ago but nothg happened. Well they didn't rewoke the jobs which I assume they could so the hope is that they will re-run the validator on all completed jobs that "failed" first time around. Failed jobs are NOT re-issued...

Very early in the morning (European time) I experienced the same but it passed and worked for 4-6 hours. This afternoon I noticed a number of completed jobs not uploaded and that the homepage was dead. I tried to reach the homepage to check status but it timed out each time. I will fire up one...

If you can figure out what the extra 30% run time is used for (real time consumption over efficient CPU time) there is room for a decent improvement ;)

Good that you figured it out. I assume that you didn't check https://www.gigabyte.com/Motherboard/B550M-AORUS-ELITE-rev-10-11-12#kf Pretty straight forward with three steps detailed. Now you know how to update a GB bios the next time and others that happens to read this thread will be pointed...

The first bios, F1 doesn't support our CPU but from F10 it does. And yes you need to flash a new bios if your MB has an older bios than the F10 since as you experienced it will not POST. You need a USB stick, FAT 32 formated, but I think that how they come from the shop and then read the manual...

And PC is back on :), well one of them. So we got 1M new jobs, enough for three days plus buffer.

@NastyHabits Then you will earn cherry pie since you are almost the only one on the team that have a "full load". I got 14 on one rig and 1 on the other so both are off untill I need to use PC power/the queue fills up whatever comes first.