Sitflyer, whaaassup? It must be cold with you now! Five others are crunching regularly but, yeah, we could do with some additional compute.
Announcement
Collapse
No announcement yet.
We're #150 @ Rosetta@Home, improving the world
Collapse
X
-
Join MURCs Distributed Computing effort for Rosetta@Home and help fight Alzheimers, Cancer, Mad Cow disease and rising oil prices.
[...]the pervading principle and abiding test of good breeding is the requirement of a substantial and patent waste of time. - Veblen
-
Tried restarting the client up after a short break. But came up with a whole crapload of errors. Looking further into it, it looks like Access Violations, so memory related. Current thinking is that the Ramdisk and Boinc don't see each other and so Boinc ends up trying to access the Ram addresses already in use by the Ramdisk. Changed the settings of Boinc to only 45% of maximum memory now, hopefully that should reduce issues from now on, and I'll keep the crunching going.
Made some minor adjustments to the BIOS settings to improve the overclock as well on the CPU. Hopefully the errors will subside and I can get back into the swing of things proper.
J1NG
Comment
-
Found the issue, it's the latest version of the Boinc Client itself that's causing issues. Rolled back to a previous version and all is well. Client is back to full swing from old config, so 4.3ghz and 9 threads. Still a while to go before I hit Zokes number though.
J1NG
Comment
-
Actually, at times, I get batches of cumputation errrors as well. May be the client but may also be the WUs. Access Violations seem to me to be a real problem. I have, on occassion, undertaken to uninstall/remove completely and re-install, with good results. I would always try the latest BOINC version becuase, so far, each one was better than the previous one (aside from taking the log-tab out and accessing it through the menu. Guy did that: No Einstein).
Good to have you back, almost! Out numbers are dwindling slowly even though MURC is thrivingJoin MURCs Distributed Computing effort for Rosetta@Home and help fight Alzheimers, Cancer, Mad Cow disease and rising oil prices.
[...]the pervading principle and abiding test of good breeding is the requirement of a substantial and patent waste of time. - Veblen
Comment
-
Modified the overclock. Remains at 4.25ghz, but boosted current and volts to various areas of the system. If it's still going to give computation errors, it's not my machine anymore. Anyway, all these errors popped up only recently when I stopped the crunching and restarted about a week ago. So I still blame the client.
Anyway, hopefully that's solved the problem now.
J1NG
Comment
-
Well, I tried. Currently getting a whole bunch of errors. And I've ran every test on this rig for the last few hours and can't find a single thing wrong with it (Memory, CPU, SSD, GPU, everything). It's only Boinc that has issues. Everything else is perfectly fine. Going to drop again for the time being until issue is found (not on my end).
J1NG
Comment
-
Don't ask me what I did, because I did nothing at all. Work Units are no longer failing. Guess it was on Boinc's end. Anyway, back in the running. Although it'll take me 18 months at least before I take Zokes spot. Plenty of time for them to return.
J1NG
Comment
-
So, what did you do??
Anyway, very nice that you've returned. We're #104 by RAC and that is not good. A sublime group like ours should be able to be top 60 easily. Unfortunaltely, some have other things to do I guess.
@Sitflyer? What's the weather like with you? You fell out, that's to bad...Join MURCs Distributed Computing effort for Rosetta@Home and help fight Alzheimers, Cancer, Mad Cow disease and rising oil prices.
[...]the pervading principle and abiding test of good breeding is the requirement of a substantial and patent waste of time. - Veblen
Comment
-
#139.
I am missing Marshmallowman as it is Winter with him now. Sifflyer has not been in for ages but it'll be getting warmer with him now. Meanwhile we have 6 millionaires.
We are just in the top 100 by RAC and we used to do way better. Anyone thinking about joining a Resurecting, that'd be great, if only for a few credits a day. Many small crunchers a large hive make.Join MURCs Distributed Computing effort for Rosetta@Home and help fight Alzheimers, Cancer, Mad Cow disease and rising oil prices.
[...]the pervading principle and abiding test of good breeding is the requirement of a substantial and patent waste of time. - Veblen
Comment
-
In the summer I do underclock my server (more). I am keeping that CPU below 50C / 120F. It is amazing what today's tech can do. My new laptop I can undervolt and clock to 1.8Ghz and let it run Rosetta on 8 threads. CPU uses about 10W (as per CPUID HWMonitor) and the fan is mostly off, only going slow occasionally. Temps are between 65C / 150F and 75C / 165F when the fan kicks in for about 15 seconds or so. This is an i7-6700HQ undervolted by 140mV. Only when I do SQL Server and heavy Excel/VBA stuff do I unleash all the power and at the same time turn Rosetta off. Thinking about working at 2.1Ghz instead of 1.8Ghz.
And yes, schedules can work very well, say, running at from 2200 to 0900 when it is cooler. That is assuming it has good powersavings and not use a lot of power even when idle. What are you running it on?Join MURCs Distributed Computing effort for Rosetta@Home and help fight Alzheimers, Cancer, Mad Cow disease and rising oil prices.
[...]the pervading principle and abiding test of good breeding is the requirement of a substantial and patent waste of time. - Veblen
Comment
-
I knew it was a bad idea to update to the latest Boinc client, it's messed up the preferences and won't fetch new work like it used to and required manual control for it to work. Grrr. Work has stalled heavily last few days. Will sort it out as soon as I can.
J1NG
Comment
Comment