Gorgon

Were at 5,189,066 so to hit 11,724,840 we need 6,537,727 over the next 13 updates or an average of 502,748 Over the last 4 updates we've averaged: ‭1,722,523‬ or 430,631 so if we keep up that rate we'll add 5,598,200 or 10,787265 total. The Czechs don't appear to be doing a final throw but Rechenkraft is so we'll have to beat 10,832,457 (434,107 average) to take them and 6th or 11,179,686 (460,818 average) to take HOCP and take 5th. So its going to be down to the wire.

Here perhaps or Here.

2.8 Million out of the gate isn't a bad start. Let’s see what the 02:00 and 03:00 updates show

Pivoted everything to the Javelin

I’ll stay on NFS till 23:45 GMT then switch everything over to Javelin unless instructed otherwise. I stopped F@H several hours ago to get another 11 threads running.

Lead is down to 728k with 3 hours to go. They’re only out producing us by 22k points per hour but that could change if the NFS servers are clearing a backlog or they’re holding bunkers

Javelin. We're pretty much un-touchable in Marathon

I believe @leadeater could get us there but only if they could bunker. Getting there without bunkering will likely require all hands on Javelin (WCG) at 20:00EST (00:00GMT)

Then we should all pivot to World Community Grid - Microbiome Immunity Project for the final Javelin throw. It would be nice if NFS' servers were working better so we could get a better understanding of where we sit so some people could start bunkering on Javelin. The event itself is over in approximately 1 day, 8 1/2 hours.

The NFS@home server appears to be not working. Their Server Stats page shows no errors other than 17,000 odd Tasks waiting for validation but I'm seeing Tasks stuck "Uploading" and jobs stuck "Ready to Report."

from The Czech National Team Forum.

I’m liking this trend: Were ahead by just over 1 million points and slowly increasing the lead. If we can maintain this we might be able to get @leadeater some breathing room to bunker for the Javelin throw.

Excellent - we're holding off the Chechs for now: So, after the Sprint is done tomorrow we should transfer all threads over to WCG MBI for the last Javelin throw then once that's done finish off on the Marathon?

Thanks Spec. Take a break and we'll see you on the other side.

@Den-Fi @LAR_Systems @LazyDev @Shlouski @porina @Eschew Can you move your system off SiDock and onto NFS@home. @leadeater says we're safe on Marathon but we need some help on the Sprint so Leadeater can bunker WCG MBI for the final Javelin throw. And don't forget to join the LinusTechTips_Team if you're signing up at NFS for the first time (like I did )

Looks like the Chechs are pushing hard to overtake on the Sprint. We'll have to keep an eye on them. It appears that we have to do the third throw on Javelin and do better than 10.8 million in order to take 5th place and pull in front of the Chechs and Rechenkraft:

You might have to change your preferences to allow the boinc-client to use more system memory. I believe the defaults are to only allow it to use 50% of the installed RAM. I had to update mine to 85% on a couple of systems to get all the threads running on NFS@home as some were stuck "Waiting for Memory" There are preferences you can set at the NFS@home site as well to only select tasks with lower memory but most of the tasks available are for the larger memory requirements.

Yup, stats broken too

Doh, Helps if I remember to join the team.

OK - Moving 131 threads from SiDock to NFS. Just waiting 10 last SiDock tasks to finish. crontabs modified to update the project every 15 minutes and submit completed tasks and fetch new ones. I presume we'll move back to the Marathon for the last day

GPUs moved off Einstein and back to F@H - That was hardly useful given the 2000 tasks left pending validation due to all the bunkering going on.

2167 here currently

Breakers, especially cheap ones, can get soft with ageing and most breakers. when running warm, will trip a little lower. I tend to run 14/3 SJ (14 gauge, three conductor Rubber (Soft) Jacketed (SJ)) extension cords I make myself with industrial grade plugs. These work well for 120V, 15A circuits we have here in North America. Much better than the crappy 18/3 extension cords with molded plugs one can buy. Though it's way over specification (1800W) for the 500-600W loads I have on them I figure there's less chance of tripping a breaker and less I^2R losses as thicker conductors=less resistance. I don't want to mess with electrical currents and want to avoid at all cost burning down the house. One of these days I'll replace the extension cords with dedicated circuits home-runned back to the panel but that's a can of worms I don't want to open just yet.

Like @Favebook said you need to set the Project Preferences at the Einstein site to only send GPU tasks then update the Projects in BOINCmanager to read the updated preferences. And check the "location" of the system to make sure it's in the same "Preference Set" (Home, School, Work ..." as the set you modified (or modify all 4 Sets). Once you click "Update" in BOINCmanager check the event log and it should show if it was successful or if not what the issue is. I have 6 systems running SiDock on the CPU and Einstein on the GPUs. It takes a little finesse to get the BOINC setup properly but once you figure it out its not too bad.

@miker07 yup, I'm coming at ya in production on Einstein ... eventually. How many tasks pending are you at? I should be at around 7.7MPPD