DC Vault 2

Thanks Gil. Just switched my clients to HPF2 from GFAM. Should hit 5 years run time before new work runs out.
 
Last edited:
I already had 5 years...I could push it to six or even seven, but I will leave the work for those badge hunting. I don't see WCG doing anything serious with badges anytime soon. They have promised a certificate for years that I personally see as worth less then the current badges. (nobody looks at the certificates provided from other projects) I'm currently 5 years and 111 days. I am betting if they do add new badge levels that 5 years will be the next stage.

WCG better do something soon. Having just finished HCC and having GFAM coming up very shortly, the loss of HPFP2 will be keeping them project light. Supposedly more projects in the pipeline, but who knows when that will be exactly.
 
I agree. WCG projects are getting thin. My goal is 5 years on the remaining projects just in case more badge levels get added.
 
I'm tapped out on 5 year goals for a while now. The only thing I had to achieve was DDDT2 which finished just shy of 5 years and of course BETA.
 
Nothing quite so high as you guys, but I luckily switched over to HPFP2 some time back, as it was the only active project I didn't have gold in, and now I'm only 40 days runtime away from the Ruby... And I agree, things are definitely getting thin!

Unfortunately, due to a bizarre computer issue I have yet to properly diagnose, I had to cut my Q9550 back to 3 cores and <90% utilization to prevent a weird freezing issue. Due to that, and the fact that my Q6600 is running MUON along side BOINC (aiming for a 20M goal), I'm only getting about 100-120 hours/day run time out of all 8 cores combined...
 
Does your computer freeze up with just running 4 BOINC WU's at a time or when you have all 4 cores tapped out regardless? Also, do you have any GPU work on that box in conjunction with CPU work?
 
It was generally freezing up with all four cores @ 100% + GPU. Stopped using the GPU, and it ran fine for a week or so, then started freezing again. Dropped it to three cores/90%, and it's running fine - except that it has frozen on me playing Borderlands the last two times playing it (last week), too, with BOINC paused. Was thinking it could be the fact that temps have gone up around here lately, but I got a new case/CPU cooler, and even at 100% load, it never exceeds 55*C on the hottest core, which shouldn't be an issue. And was thinking it might be the PSU, but I just bought a new Corsair TX750 at the same time I bought the new case, so 750W should be handling a non-overclocked Q9550 and a single 9600GSO without breaking a sweat!

So now I'm down to either 1) memory (which was the problem on my Q6600 system), 2) motherboard old and dying (this is a crappy MB that can't OC at all), or 3) something worse. I want to upgrade this system anyways, so I'm just trying to limp it along until Haswell comes out and I can upgrade to it or buy a used SB/IB system for cheap from someone on the DC forums (that's how I got my Q6600)...

Of course, I say that now, but as soon as I do, I'm sure I'll be looking to take all my extra parts (mostly this rig) to rebuild another DC machine for the farm. After all, my 9 year old daughter needs her own folding/BOINC machine, right? :D
 
I've read many forum posts that are similar to yours that wound up being one bad component that randomly had trouble. Keep in mind it could be one core in the CPU causing the issue. Since the work units bounce between cores, one core might actually get much higher then the others at some point triggering a weak spot or flaw in the core. Thus the freeze or crash. You have a very long road ahead of you in figuring that out. Best wishes and luck to you.
 
I added another phone to my Android/NativeBOINC Pharm. LG Optimus V
It isn't a power house by any means, but it works. So that makes 3 phones with 600MHz single core processors being put to work rather then being thrown into an ecyclers hands. My 4th phone is a dual core 1GHz that I still use, so it only works part time.

I am glad to see that we now have 3 members using NativeBOINC. Myself, CaptainUnlikely, and now tiderfish. Keep up the good work Commandos!
 
Last edited:
I'm almost beginning to wonder if a small Android DC farm is worthwhile. You can pick up small quad core boards fairly cheaply, extrapolating from my 1.6GHz tablet they would do roughly 1500RAC on Enigma for around 10W power draw. How do you guys think that compares to more traditional crunchers in terms of performance/watt?

I'm adding a quad core phone to my NativeBOINC arsenal soon, a Tegra 3 based LG Optimus 4X. As soon as I can figure out how to type on a touchscreen phone, I'll retire my trusty Desire Z in favour of this and it'll crunch roughly 9 hours a day as my Desire Z does at the minute. Although, it does have a reputation for getting hot, so I anticipate much messing with custom kernels etc before I'm satisfied it won't overheat during the night.
 
It all depend on the project and the ARM processor. The older ones have to have software emulation designed into the app for some of the projects to work on them. The newer processors work quite nicely. However, I haven't done/seen enough research and comparisons to actually decide it investing in them specifically for BOINC is worth it over other options. I don't buy the ARM products for that purpose. I just repurpose them after the fact. :)
 
S4 Pro in a Nexus 4 is doing swimingly in SETI. Estimates 40 or more hours but only takes between 4-10 in reality.
 
Kai, don't forget to bring Rosetta to our team as well. ;) Glad you are aboard. Hopefully NativeBOINC updates soon so that it shows a fourth member using Android.

And thanks for the instant bump down in the rankings at Einstein. :D
 
Haha, oops! I swear I didn't mean to. What a way to make a first impression, right?

Oh yeah! Rosetta! Almost forgot about that little gem. Not quite as "cool" as FAH, but no less important!
 
Yeah...but it still helps our push in the DC-Vault. http://www.dc-vault.com/ This month a few of us focused on Correlizer. Next month, I plan on focusing on Rosetta. It has been a while since I hit it [H].

On a side note, my 1 million point goal at Sudoku (non-Vault) is 93% complete.
 
Correlizer seems to be down, my laptop has crunched through all downloaded work and isn't getting any more or uploading the finished tasks, I also can't access the web site right now. I've set it back onto Climate Prediction for the time being.
My phone seems to overheat quite badly when running NativeBOINC, to the point where it stopped charging properly until I shut down the client and le tit cool back down, so that looks like a no-go sadly.
 
Wow...never seen NativeBOINC do that on a phone yet.

Correlizer has been down a few days. I have no idea why, so my backup projects are just getting a little extra attention.

This would be a good time to have the Commandos decide whether we have a unified push this month or not. I plan on hitting Rosetta@home. I also have a few cores on WCG helping with finishing HPFP2 up.

And, welcome Kai to the Commandos. xENo, good job on Rosetta.
 
Thanks for the welcome!

I agree with a concentrated push, and Rosetta seems like a great candidate.
 
Well, I just got my Ruby badge in HPFP2, and am going for Ruby in GFAM before it finishes up. That should take me into mid-June, but as soon as I hit that, I'll be happy to jump on whatever the June project is...
 
Wow...never seen NativeBOINC do that on a phone yet.

The phone I have now is known for having some heat issues. Once I've made sure it's not about to suddenly die on me, I'll get a custom kernel & ROM on there at the weekend and play about with the settings, maybe lock the CPU to a lower speed and see if I can get it running without overheating.

This would be a good time to have the Commandos decide whether we have a unified push this month or not. I plan on hitting Rosetta@home. I also have a few cores on WCG helping with finishing HPFP2 up.

I'm happy to hit Rosetta with you guys then, still haven't got much to throw at it (only one core, can't guarantee the other cores at my disposal will meet deadlines) but I hope to finish my build this month so I'll hopefully be able to hit July's project a little [H]arder.
 
For the WCG'ers... "We&#8217;re happy to announce that a new project from the Ontario Cancer Institute is in the works! Stay tuned for the reveal of the project name and description later this summer! There are several other projects in the pipeline, so rest assured that you will still be able to contribute to important research." http://www.worldcommunitygrid.org/about_us/viewNewsArticle.do?articleId=302

That is the same group that handled HCC. Also, they mention having a few more surprises in store for us. Thus, there will probably be a WCG push in the near future. With new projects means new badges. A push there would appeal to all except point whores and potentially GPU crunchers. :D (GPU is still unknown since that group were the ones who brought GPU to WCG.)
 
For those who have updated to the latest alpha client 7.1.1, be careful raising your cache. If you set projects to No New Work, the manager doesn't fully honor your settings. I tried to load up some WCG work on a few machines and started getting work for projects that had been set to NNW for quite some time.

WCG's HPFP2 has about 11 days of new work left and GFAM has ~4 days of new work yet. If you haven't gotten your goals yet, better start loading your cache.

http://s137.photobucket.com/user/Sekerob/media/WCGYearsPi1Project.png.html

I have started attaching systems to Rosetta. I will be full swing when my WCG caches empty.
 
Damn, looks like I'll miss Ruby on Gfam - even if I fill my queue this weekend, I won't have enough time to meet the deadlines...

Edit: screw it, I'm gonna give it a shot... The current deadline has moved from 6/3 to 6/9, so I can get ~10 days x 4 cores x 2 CPUs = 80 days. And I only need 75 days time for ruby. Gonna cut it close, but I think I can get it! Time to go try to boost the OC on the Q6600 a bit more :D
 
Last edited:
SETI Enhanced V7 now out. Looks like a big performance boost with OpenCL and Cuda and it uses less CPU resources too! About 700 seconds with a 7970 GHz, and about 800 seconds or so for a superclocked Titan. Not bad!
 
Climate Prediction launched their new home page. Looks nice but is kind of clunky for BOINC user navigation. Not that the old one was great to begin with.

I also just had a Q8200 die tonight. The motherboard has had tons of swollen capacitors for 4 years and tonight it stopped detecting SATA drives. The IDE DVD burner still detects, so I may slap an old IDE hard drive in it to get it up and running. I haven't decided yet. I've been thinking of finding a cheap socket 775 motherboard, but I would like to find one with multiple PCI Express slots for running multiple GPU cards. I have plenty of DDR2 memory, but I could scrape up some DDR3 if I had to.

Update 2...After multiple restart attempts and CMOS power drains etc... The Q8200 lives once more. I think this time I will use the life of it to find a replacement board...
 
Last edited:
Wow, I hope nothing explodes while you look for another board. Would be quite the hot mess for sure!
 
Badge Hunters.... PrimeGrid's new sub-project SR5 goes live June 14th. It is CPU based.
 
Just got word from themacguy our shared farm is going offline today. We'll still have 24 cores between us crunching at home until September.

I'm getting close to my 5 year goal in HPF2. Should reach it later this week. My plan is to run DSFL, HFCC, SN2S and FAAH until they all reach 5 years. Should take a couple months.

I started moving cores over to DSFL this afternoon.
 
Collatz got zapped by electrical storm. Make sure if this is one of your projects that you have a backup project ready.
 
Kai, don't forget to also attach to WUProp. It is a non-cpu intensive BOINC app that runs alongside your other projects. Think free points and badges. It is also a good reference points for how different projects perform and what to expect.

It appears Rosetta is taking GPU possibilities more seriously. Cross your fingers. http://boinc.bakerlab.org/rosetta/forum_thread.php?id=6185&nowrap=true#75022
 
Last edited:
FYI, GFAM ran out of work on BOINC yesterday. I have 54 days to my Ruby goal with 67 days queued, so if I can hit the deadlines, I will hit my goal with 10 days run time to spare...
 
Finally got my Mac Pro 2 x X5675 running without the fans sounding like a jet engine. Currently attached to Rosetta at the moment. Before the 10.8.4 update the SMC wouldn't let the temp go above 45C, thus the fans were at max, and now it lets it get a little higher while letting the fans sit at a reasonable level. Should have 2 x 6344's up and running by next month hopefully. Fun stuff!
 
Awesome to hear. My output should increase within the following week. I had to clear my cache of WCG HPFP2 work, get my fathers internet going again (yeah phone company swore it wasn't them...guess what?.....), and am switching laptops this weekend (hopefully) from a single core sempron to a dual core. My in-laws are having a yard sale tomorrow as well, so hopefully I can get rid of some of these old HT P4's and single cores that are collecting dust in my office. They are borged for now, but I had to shut them off due to hot weather.
 
Back
Top