• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

DC Vault 2

Awesome to see you got internet again DooKey. I had a few systems go down unavoidable recently too which caused me missing my goal at WCG for CFSW project. I still have a few more goals at that project before readjusting resources, but am adding more hardware and upgrading some borged systems. Hopefully, my PPD will get a good boost soon.

We are currently ranked 96th for points at FreeHAL, 90th for WUProp, and 8th at OProject for points. Each of these projects have non-CPU intensive apps that can be run alongside other BOINC projects or even alongside other non-BOINC projects. I think [H] could do a lot more in these projects while supporting their favorite ones. I would however warn people that OProject is still working out the kinks, so their NCI app ALX still has a few minor issues.
 
I also forgot to point out Quake Catcher Network for those running Macs. It is an NCI BOINC app as well and it is monitored by Stanford. Some IBM's also work without any additional hardware. Otherwise, you would need to purchase one of their sensors to get the app to work. Without the sensor, it would just run a their demo screen saver. We don't have a lot of members at that project and figured some of the Commandos might have macs laying around.
 
For those wanting to do distributive computing on their Android devices, you should give nativeBOINC a try. http://nativeboinc.org/site/news I have been using it for several months now and have found it to be pretty impressive. Previous fears were that the phones or tablets would get too hot or take too long to finish work, but this is not true. It will drain the battery though like any other app. Currently there are only a handful of projects that support Android + ARM processors. PrimeGrid, enigma, WUProp, OProject, Subsetsum, and Milkyway are the current list that I know of. I have it running on my Bionic Motorola phone and run it at night while I'm sleeping.

http://milkyway.cs.rpi.edu/milkyway/show_host_detail.php?hostid=450170
http://www.enigmaathome.net/show_host_detail.php?hostid=80520
http://volunteer.cs.und.edu/subset_sum/show_host_detail.php?hostid=2171
http://wuprop.boinc-af.org/show_host_detail.php?hostid=36097
http://oproject.goldbach.pl/show_host_detail.php?hostid=651
http://www.primegrid.com/show_host_detail.php?hostid=261114
 
Well life has bee a little too busy lately so I havent got around to posting much so sorry for putting so many random thoughts from the past couple weeks in one post. :)

1. @Razor - Good to see you busting it on SoB man.
2. @dmaine - Congrats on breaking the 100 million mark in total BOINC points! A job well done!
3. @Gil - Thanks for the heads up on the non-CPU intensive projects. I need to read about these projects a little more but if they run nicely with other CPU intensive BOINC projects i wouldn't mind trying out a couple. And sweet find on the Andriod BOINC crunching. Thats [H] for sure!

4. DC Vault News - It looks like LHC@home was recently added to the DC Vault. I havent been able to find any discussion about adding this project (or even a request for it to be added) to the vault and when I searched the web i found more than one seemingly active DC project page based on LHC research so Im not 100% sure on this yet. If anyone has any indo let me know. Otherwise I might ask on the DC Vault forum.

5. Props to the WCG crew. We are slowly progressing an upward trend. I'm still planning on trying to join the Christmas race again in Dec. I hope you are all adding new boxen for the race! :D Also in WCG News - The Help Cure Muscular Dystrophy - Phase 2 and Computing for Sustainable Water projects are concluding. More info at http://www.worldcommunitygrid.org.

6. Anyone want to get a small project push together for like the second half of Oct or something. Im thinking that it might be a good idea to do some more Muon1 as OCAU is coming up quick behind us and we should still be able to get another spot in front of us before the Christmas Race if we make a good push. Thoughts?
 
metallicafan said:
I think I can speak for all of the Commandos when I congratulate DooKey on taking the #1 spot in total BOINC points with over 240M points! tr0ach has held this position for quite some time and also deserves a pat on the back but DooKey has been churning out better than 1M PPD for quite some time now. Well done sir! We are glad to be crunching with you!
Cheers!

http://stats.free-dc.org/stats.php?page=userteamcomb&team=[H]ard|OCP
Click to expand...

Congrats DooKey!
...
...
...
...
Hey, I'm only down to #22! Might still have a shot at top 10... :D Where'd I put that damn thermal paste...
 
metallicafan said:
4. DC Vault News - It looks like LHC@home was recently added to the DC Vault. I havent been able to find any discussion about adding this project (or even a request for it to be added) to the vault and when I searched the web i found more than one seemingly active DC project page based on LHC research so Im not 100% sure on this yet. If anyone has any indo let me know. Otherwise I might ask on the DC Vault forum.
Click to expand...

It was in the Vault way back in the day, but got dropped due to inactivity/lack of work. Usually when a project gets dropped they say "we'll add it back if/when..." So those conditions were likely met recently.

And we're currently 434th in LHC for 8807 vault points so that might be a good candidate to toss some resources at.
 
alan2308 said:
It was in the Vault way back in the day, but got dropped due to inactivity/lack of work. Usually when a project gets dropped they say "we'll add it back if/when..." So those conditions were likely met recently.

And we're currently 434th in LHC for 8807 vault points so that might be a good candidate to toss some resources at.
Click to expand...

I'll throw my 3930K at it tonight to see how it runs.
 
Quick vault question - I have a chance to score a free AM3+ mobo, and was wondering if slapping an FX-8150 in it for 159 from Newegg is worth it, or if I should use the money towards an IB i5/i7? Seems like it would make for a good 8 WU at a time box for under 200, but I wanted to see if anyone had hands on experience with AMD and vault projects recently... And if so, will 4x4GB ram be enough to push all 8 WU at a time, or should I go 32GB (making it a lot more expensive)?
 
16GB should be ample. An FX-8150 will definitely be worth it, but depending on how concerned you are with power consumption you may want to move to a SB/IVB and sell the AM3+ MB.
 
Jathanis said:
Quick vault question - I have a chance to score a free AM3+ mobo, and was wondering if slapping an FX-8150 in it for 159 from Newegg is worth it, or if I should use the money towards an IB i5/i7? Seems like it would make for a good 8 WU at a time box for under 200, but I wanted to see if anyone had hands on experience with AMD and vault projects recently... And if so, will 4x4GB ram be enough to push all 8 WU at a time, or should I go 32GB (making it a lot more expensive)?
Click to expand...

In general, 16GB should be fine. I don't think any of the BOINC based projects in the vault have any work needing more then 1GB of ram per work unit. You didn't mention how many GPU's you would plan on running either. The GPU's wouldn't normally be a big deal unless you told BOINC to leave applications in memory. Then when work pauses, the GPU's would also go to memory, but I don't know if it would sit in V-Ram, Ram, or get dumped to Virtual Memory. As of right now, the vault doesn't have any of the NCI apps included. In that case, you could easily consume more then 16GB of ram. Especially if you set up a few virtual machines to monopolize on such projects.
 
jathanis - I admittedly know nothing about AMD's newer processors but it seems like if you are going to take advantage of the free AM3+ board as a dedicated folder than 16GB should be plenty for about any DC project. Even the WCG's Clean Enegry Project - Phase 2 "only" needs about 1GB of RAM. If you were using the machine for VM's of something you might need more but 16GB should do it for crunching.

Also a chip that runs 8 threads will be great for chasing WCG badges. ;)
 
mmmmmdonuts said:
16GB should be ample. An FX-8150 will definitely be worth it, but depending on how concerned you are with power consumption you may want to move to a SB/IVB and sell the AM3+ MB.
Click to expand...
I guess then the big question would be what would produce more, the 8 cores of the FX-8150 or the 4 cores of an i7 IB - and would the energy savings offset the extra cost of the more expensive Intel processor & matching motherboard?

Gilthanis said:
You didn't mention how many GPU's you would plan on running either. The GPU's wouldn't normally be a big deal unless you told BOINC to leave applications in memory.
Click to expand...
Well, my first goal is to get the q9550 in the gaming rig in my sig able to actually do WUs again without overheating. I think the old heatsink I've been using has finally "died" - it's the old kind without a backplate, and the plastic tabs won't hold it tight enough to keep contact with the CPU anymore. Right now just typing this with ambient temps at 75*F, I'm idling at 47*C on my hottest core. 2-3 hours of BOINC and I get reboots or throttling... Problem is I also have it in a crappy case without a cutout on the MB tray, so to replace the heatsink, I have to remove the entire Mobo, several drives, etc... So I want to get a new case with better airflow and rebuild my gaming rig to get 1 box running. After that I want to get the HTPC q6600 running, but need a new PSU & HDD. Then I'll be back to where I was before I took a break... Both easy fixes, now that I have some money.

Then, I planned on building a new rig for myself after Christmas, looking at a i5-2500K, i7-2600K or i7-3770K. Then I'll move the q9550 over to my daughter's desk, and have 3 quads for crunching. Then this deal for the AMD board came up, so I was considering making that into a fourth (headless) box, just the board, CPU and memory hooked up to a PSU, no major GPU at first. I figure that gives me 8 more cores for under $400. So after I figure all THAT out, then I'll look at adding GPU power to all of it! I'm sure the old nVidia 9 series cards just don't do much for the power draw these days...

Problem I have, I guess, is too many projects to get back into crunching for the Commandos, not enough time or money until after Christmas!
 
So based on this in-depth post, I think I'm going to skip the FX-8150 idea, and maybe pick up a old X6 1090T I found online for $115 + shipping: http://www.overclock.net/t/1143670/fx-8120-boinc-benchmarks-real-world-stuff - Two less cores, but better performance at half the cost. Can't beat under $250 + old parts to add 6 cores to the cause!

On a related note, I pulled the heatsink & reseated it on my gaming rig, and am again running some BOINC! Only 3 cores for about 1200 ppd on 2 WCG projects I've never done, and staying below 70*C under load, but once I add a couple bronze 15 day badges, I'll start jumping on some other projects. And I have a new PSU on the way, so hopefully the q6600/gtx9800 rig will be back online in a few weeks as well! :D
 
metallicafan said:
Sweet. Im on it. Looks like this might shake up scoring on the WCG a bit. :) Thanks for the heads up.
Click to expand...

Just a heads up again, those looking to do GPU for the HCC app, they don't currently have a way to choose to just do GPU and not the CPU version of it. Supposedly they are working on that. Until then, we just have to do both if we want them.

And to those whom don't keep up with WCG, the HCC app running on GPU will reserve an entire core for the app while it runs. So unlike other GPU projects, it doesn't leave the majority of the core to do other things.
 
My name is Tobitello and I approve of kenlo's [H]ard work. :cool:
 
Thanks for the F@H help everyone, much obliged. :D
Any concerted efforts lately or just pick and crunch? I am avoiding GPU projects for now, electric bill is gonna be through the roof. :eek:
 
There are still 6 or 7 spots we could pick up in Sztaki if folks would like to run that for a few weeks.
 
Gilthanis said:
Just a heads up again, those looking to do GPU for the HCC app, they don't currently have a way to choose to just do GPU and not the CPU version of it. Supposedly they are working on that. Until then, we just have to do both if we want them.

And to those whom don't keep up with WCG, the HCC app running on GPU will reserve an entire core for the app while it runs. So unlike other GPU projects, it doesn't leave the majority of the core to do other things.
Click to expand...
The HCC project seems to be running pretty well on my Nvidia 550Ti. I pretty much have to pause it when I use the machine because of the demand it places on the system but other than that it seems pretty stable. @Gilthanis - How long are your WUs taking to complete on your cards? Im getting about 6-7mins depending on the WU on the 550Ti.


Razor_FX_II said:
Congrats kenlo on 3rd place over all.:D
Click to expand...
Congrats kenlo! Nice works sir! Especially on Distrrtgen!
 
metallicafan said:
The HCC project seems to be running pretty well on my Nvidia 550Ti. I pretty much have to pause it when I use the machine because of the demand it places on the system but other than that it seems pretty stable. @Gilthanis - How long are your WUs taking to complete on your cards? Im getting about 6-7mins depending on the WU on the 550Ti.



Congrats kenlo! Nice works sir! Especially on Distrrtgen!
Click to expand...

Quite honestly, I'm not sure how long they are taking. My cards are working hard elsewhere at the moment and most of the HCC GPU work is on my borged machines. I will try and look shortly.
 
On the 660TI, I think it is between 4 and 5 minutes but that is borged. The 640 is still down and in a borged machine. A borged AMD Radeon HD 6550D (A8-3800 APU) 512MB APU is processing also between 3 and 4 minutes. General concensus right now is that AMD is faster then nVidia. Once my home PC's are done with their other work, they may chip in. However, my fastest cards in my home are the GT430 (which I have 3 of). I also have a 520 (temporarilly while fixing a family members PC), and then a 4350 and some 210's and 8400's that don't seem capable of running any HCC yet but are available if they ever figure it out.

Edit: Also check out this thread. Specifically the post by OldChap. http://www.worldcommunitygrid.org/forums/wcg/viewthread_thread,33962_offset,140 He has a breakdown on some various cards he is using.
 
Last edited:
For those interested in OProject, they are limiting the ALX app to one per system rather then per core or thread like FreeHAL does. I also have found that by running just ALX, I have a much higher RAC then running their other apps too. To get around the one ALX at a time, I recommend using virtual machines.
 
Well, couldn't get my video card to grab any HCC WU, so I went to the forums and found that all my video cards are too old to use... Any suggestions for current BOINC projects I could throw an ATI 4850 at?
 
Collatz, MOO!, Primegrid, and (I can't remember if that card was dual precision or not) Milkyway. You may even be able to run it at POEM, Albert or Donate@home. SETI BETA might also work.
 
Gilthanis said:
Collatz, MOO!, Primegrid, and (I can't remember if that card was dual precision or not) Milkyway. You may even be able to run it at POEM, Albert or Donate@home. SETI BETA might also work.
Click to expand...

Thanks for the input - I've already gotten to 10M on Milkyway and 5M on Collatz with this card in the past, so I was looking for something new... Since I already have an account on Poem, I think I'll try that for starters. :D
 
Jathanis said:
Thanks for the input - I've already gotten to 10M on Milkyway and 5M on Collatz with this card in the past, so I was looking for something new... Since I already have an account on Poem, I think I'll try that for starters. :D
Click to expand...

I don't know the video ram on your specific card, but make sure you maximize it with an app_info.xml file. I am running two at a time on an nVidia GT 520 and doing much better then running one unit at a time. My 210 currently running POEM only has 512MB, so I'm only running one on it. I'm so spread out right now, I'm not sure if my 210 cards with 1GB will do OK or not with multiples. Also, keep in mind that many find better results dedicating a full CPU thread/core to feed the GPU there.
 
Gilthanis said:
I don't know the video ram on your specific card, but make sure you maximize it with an app_info.xml file. I am running two at a time on an nVidia GT 520 and doing much better then running one unit at a time. My 210 currently running POEM only has 512MB, so I'm only running one on it. I'm so spread out right now, I'm not sure if my 210 cards with 1GB will do OK or not with multiples. Also, keep in mind that many find better results dedicating a full CPU thread/core to feed the GPU there.
Click to expand...

Well, I'm giving up for tonight, as I've tried tweaking aboout everything I remember to no avail... I couldn't get any GPU work units with Poem, even after making sure to change my preferences on the Poem project site to allow ATI WUs. I then tried Primegrid, and got a handful of units right away, but they all errored out. Reading their forums and FAQ, I realized that I'm running old drivers, and that either the APP drivers or the ATI Stream APK (only available for Vista/7, and I'm running XP) is required, but not until AFTER updating my drivers to the latest Catalyst drivers... Now the options for GPU on the BOINC activity menu have disappeared, too, and nothing GPU will run! :mad:

Perhaps I'll look at Donate or Albert tomorrow, or even go do another 5M on Collatz... And FWIW, it's a 512MB 4850.
 
Keep in mind the version of BOINC also matters to some projects now as well. Some require version 7.X..X and some don't. (Albert is one of those) You will need to install the latest Catalyst package from AMD which includes everything needed these days. Both AMD and nVidia are dropping XP support. Some people have had luck installing the new Catalyst over the old rather then a clean install. You may want to try that. Also, make sure you aren't running BOINC as a service. This will disable any option of using the GPU.
 
Jathanis said:
Well, I'm giving up for tonight, as I've tried tweaking aboout everything I remember to no avail... I couldn't get any GPU work units with Poem, even after making sure to change my preferences on the Poem project site to allow ATI WUs. I then tried Primegrid, and got a handful of units right away, but they all errored out. Reading their forums and FAQ, I realized that I'm running old drivers, and that either the APP drivers or the ATI Stream APK (only available for Vista/7, and I'm running XP) is required, but not until AFTER updating my drivers to the latest Catalyst drivers... Now the options for GPU on the BOINC activity menu have disappeared, too, and nothing GPU will run! :mad:

Perhaps I'll look at Donate or Albert tomorrow, or even go do another 5M on Collatz... And FWIW, it's a 512MB 4850.
Click to expand...

Any luck with the 4800 series cards Jathanis? I tried several different projects a few weeks ago on my 3870 card but could manage to find on that still supported this card. not complaining . . . I know its an old card and graphics tech is moving fast. Just hopeful you found a project to put its crunching power towards.
 
Gilthanis said:
metallicafan, my nVidia GT 430 in my Q8200 is taking about 18 to 20 minutes on HCC GPU work units. I have not yet ran any that I've noticed on my i7 920 that has 2 GT 430's in it. I will update you if I see a difference.

Also, take a look at this sticky post with some official numbers: https://secure.worldcommunitygrid.org/forums/wcg/viewthread_thread,34055
Click to expand...

Thanks for the results and the link. Looks like AMD cards are still crunching faster than Nvidia cards even in WCG. Any [H]Users running HCC on an AMD card care to share their experiences?
 
My experience with GPU projects is that if it is OpenCL, then AMD wins. I hear that future nVidia cards may get better with this, but we always hear things like that.
 
You must log in or register to reply here.
Back
Top