BOINC Community Board

[Ithanul]

On 3/17/2020 at 5:27 PM, MrFinn741 said:

In lack of FAH WUs i tried running asteroids at home, but all GPU tasks error out on start. Is there any known problems?

I have two GPUs, a 1660Ti and a mining 1070, can I somehow choose what GPU Boinc uses, the 1070 shows up as the first GPU so maybe Boinc doesn't like such an odd card?

Is there anywhere I can check whats wrong, logs or other messages, it doesn't show anything in the Stderr output on the website just says Computation error.

System is stable I have been running FAH for a week with small overclocks and no change when I removed them.

Hmmm, I run an older 1080Ti on occasion with asteroids at home with just basic setup.

 

In the BOINC client, you can get a log window to open by either  clicking on -> Tools -> Event Log, or the shortcut -> Ctrl+Shift+E

 

Otherwise, may have to dig over on their forum board to see if there are any current issues with driver or specific cards.

[42thgamer]

Joined the team!

[MrFinn741]

On 3/21/2020 at 8:54 PM, Ithanul said:

Hmmm, I run an older 1080Ti on occasion with asteroids at home with just basic setup.

 

In the BOINC client, you can get a log window to open by either  clicking on -> Tools -> Event Log, or the shortcut -> Ctrl+Shift+E

 

Otherwise, may have to dig over on their forum board to see if there are any current issues with driver or specific cards.

I found the log but no error messages. I was on their forum and it seems to be a problem with the 16xx series GPU:s and maybe the rtx series. I also found out how to get multiple GPU:s running with Boinc, now I just have to study some more config file writing to force a certain project to a certain GPU. I am running GPUGRID, it just works, in the mean time as I haven't had time to investigate the config and F@H is again dry on WU:s.

[AlpacaTechTips]

Joined the team. Folding Rosetta@home project.

[Will Treaty]

Shouldn't we do a event for Rosetta@home project? They are working on COVID-19 as well and due to their expiration dates (much more than F@H) we can call the people who uses low-end pcs and able to support a different project.

[Ithanul]

11 hours ago, MrFinn741 said:

I found the log but no error messages. I was on their forum and it seems to be a problem with the 16xx series GPU:s and maybe the rtx series. I also found out how to get multiple GPU:s running with Boinc, now I just have to study some more config file writing to force a certain project to a certain GPU. I am running GPUGRID, it just works, in the mean time as I haven't had time to investigate the config and F@H is again dry on WU:s.

Also, I just got this through BOINC from them.  Maybe with the new WUs there will be less issues with 16xx series GPUs.

 

Quote

Asteroids@home: New CUDA application for Windows x64 Release We are proud to present you the new CUDA application that was released by our team yesterday. What's improved? - We've Added support for latest Compute Capabilities (CC). Application was built with latest CUDA SDK v10.2 adding support for latest NVIDIA devices with Compute Capabilities (CC). This build supports CC 3.0, 3.5, 3.7, 5.0, 5.2, 6.0, 6.1, 7.0, & 7.5 from Kepler, Maxwell, Pascal, Volta & Turing architectures and their GeForce, Quadro, NVS and Tesla series, but does NOT supports Tegra and Jetson SoC Series devices (CC 3.2, 5.3, 6.2 & 7.2) which are not subject of interest for this project. - Fixed issue with special Light curves which has more than 1000 points. There is no x86 (32bit) version. CUDA 10.2 supports x64 (64bits) operating systems only. Side effects that could be observed because of the update. If your client shows unusual "Remaining time" for the tasks you can follow these steps: 1) Set "No new tasks" from "Project" tab for "Asteroids@Home" project. 2) Then you need to finish your recently downloaded tasks. 3) When all are done and reported to the server please reset the project with 'Reset project' button at Project tab. We also suggest to use the latest stable version 7.14.0 of BOINC client. And last but not least we strongly suggest users, who has CUDA enabled devices to upgrade their drivers to the latest one. For today it is 442.74 - WHQL. You will find it here: https://www.nvidia.com/Download/driverResults.aspx/158759/en-us For legacy drivers you will find drivers here: https://www.geforce.com/drivers/beta-legacy Despite new CUDA application is release we will continue to support older devices with older drivers by the previous CUDA55 application. I would also like to thank Georgi Vidinski who made this release possible. Radim Vančo (FoxKyong)

 

 

14 hours ago, 42thgamer said:

Joined the team!

 

5 hours ago, AlpacaTechTips said:

Joined the team. Folding Rosetta@home project.

Welcome to the team.  

 

1 hour ago, Will Treaty said:

Shouldn't we do a event for Rosetta@home project? They are working on COVID-19 as well and due to their expiration dates (much more than F@H) we can call the people who uses low-end pcs and able to support a different project.

There is plans to run a mini BOINC event, but that has been push off to not overlap with the current F@H event.

The plan was to run a mini event as a warm up for the upcoming Pentathlon 2020 for BOINC.

[MrFinn741]

4 hours ago, Ithanul said:

Also, I just got this through BOINC from them.  Maybe with the new WUs there will be less issues with 16xx series GPUs.

I saw that, hopefully it will work, it should as it says support for Turing. I haven't tested yet but will as soon as the GPUGRID tasks run out

[rkv_2401]

Hey guys, I can confirm that Rosetta@Home is still sending out COVID-19 WUs.

 

I have another one in the queue that I got today(24th of March), with "COVID-19" in the name.

[Wouterrinos]

Hi, I want to join the Rosseta@home project, but it keeps saying my email adres isn't valid. I joined an other project to see if it would say the same there, it didn't. PrimeGrid is running now. I tried making an account on the Rosseta@home website and I get the same message. Maybe it is beacuse my email adres has points in it, but that would be pretty stupid.

[MrFinn741]

2 hours ago, MrFinn741 said:

I saw that, hopefully it will work, it should as it says support for Turing. I haven't tested yet but will as soon as the GPUGRID tasks run out

Tested and no more errors only completed tasks

[Flatronic]

14 hours ago, AlpacaTechTips said:

Joined the team. Folding Rosetta@home project.

I gave them all my CPU power due to the fact they doing those calculation not with GPU's

[AlpacaTechTips]

2 hours ago, Wouterrinos said:

Hi, I want to join the Rosseta@home project, but it keeps saying my email adres isn't valid. I joined an other project to see if it would say the same there, it didn't. PrimeGrid is running now. I tried making an account on the Rosseta@home website and I get the same message. Maybe it is beacuse my email adres has points in it, but that would be pretty stupid.

you don't have to confirm your email address - or at least i did not, so you can put in a fake one if you like. just remember the email, the name and password so you can log in when you need to.

[AlpacaTechTips]

9 minutes ago, Flatronic said:

I gave them all my CPU power due to the fact they doing those calculation not with GPU's

I did too. Seemed pretty good balance to run CPU for Rosetta and GPU for F@h.

[Wouterrinos]

20 minutes ago, AlpacaTechTips said:

you don't have to confirm your email address - or at least i did not, so you can put in a fake one if you like. just remember the email, the name and password so you can log in when you need to.

Yeah, no thanks. I have to remember plenty of stuff already, I'm not going to make up a fake email because their system doesn't allow me to use my actual one. PrimeGrid it is (and F@H).

[El_Psy_CoNGR00]

hey every one. i have come over from the f@h thread to introduce myself and let you know i have now joined the boinc team. i was struggling for wu's so i decided i would fold r@h whilst folding with f@h alongside  on my gpu.  

[Crabbular]

Hey I also have started R@H on my CPU with F@H on the GPU 

[JordNL]

On 3/24/2020 at 10:13 AM, Wouterrinos said:

Hi, I want to join the Rosseta@home project, but it keeps saying my email adres isn't valid. I joined an other project to see if it would say the same there, it didn't. PrimeGrid is running now. I tried making an account on the Rosseta@home website and I get the same message. Maybe it is beacuse my email adres has points in it, but that would be pretty stupid.

No, my email address has dots in it as well, and it works fine, so that ain't it. What Rosetta does though is check all incoming email addresses against the stopforumspam filter which may hold your email domain. Protonmail is one of the mail addresses that cannot be used there. You also can't use a VPN when signing up.

 

But a very simple solution if you haven't tried yet, is to make an account via BOINC Manager->Add project wizard. That doesn't get the rigorous check, as far as I know.

[Crabbular]

I got one of those dodgy x79 AliExpress motherboards and an old Xeon E5-2648L for more BOINC, was going to use it as a virtualised router and pi hole etc. but that can wait.
 

I think the memory is running quad-channel, but I actually have no idea 

[JordNL]

A couple of answers from earlier posts, so far they haven't been answered yet. I haven't read the complete thread. But any questions you have and don't think have been answered yet, please post them for me.

 

Who am I then? I'm Jord (also know as Ageless here and there), from the BOINC forums. Branching out a little. I won't be able to answer all, like what projects best run on the CPU and what on the GPU or what hardware is best to use as these are a personal decision. For instance, I run 24/7 on a couple of Android devices I have lying around, they do Seti, Asteroids, Einstein and Universe. My 3900X runs Rosetta at the moment.

 

1. Credits difference between projects: BOINC is open source, meaning that anyone can download the source code for the client and the back-end and adjust it to their liking and run it that way. This goes for the credit situation as well. Although BOINC does have a credit function built in, most projects disable this and set up static credit per task. With as much credit as their users want, this to lure others over to run their tasks and get more credit. Since there's no monetary function to this credit (unless you run Gridcoin), the only thing you can do with it is boast to others how much credit you gained in a day.

 

2 "Aborting a WU is mostly seen in the community as a no, no." Not if you want to get rid of a lot of tasks quickly. Resetting the project here is quicker, but then all the tasks you had on your system are just deleted with no word back to the project and they need to time out before they're reissued. Which at some projects can take months. That's seen as a big no no. Aborting tasks at least deletes them from your system but also reports them back to the server as being aborted, which means they can be reissued immediately.

 

3. Uploads vs reporting. Uploads happen automatically and immediately upon task finish, as long as your Network is enabled. Uploading is just moving bits of data from your hard drive to a hard drive at the project.
Then the task has to be reported to the database, with whatever outcome it had, good, bad, aborted, out of time, etc. These can stack up and BOINC does this stacking because reporting uses server overhead (use of memory and CPU). Reporting 1 task takes as much overhead on the server as reporting 100 tasks does. So projects rather have you report as many as possible at the same time. Because your computer isn't the only one making contact with the project, there are thousands of others there as well, all with that same overhead. You can't force an upload to happen with the Update button, but you can force a report to happen with that button.

 

4. Not all projects run the same BOINC server version, back-end or front-end.
Einstein developed a Drupal front-end that runs on the normal BOINC back-end, but then they stopped updating that so it's all a bit old.
World Community Grid was already a Grid computing entity before they joined forces with BOINC. This is why they have their own points distribution for instance, with 1 BOINC credit being equal to 7 WCG points. WCG is an iniative of IBM.

Most projects have updated their back-end to a newer BOINC server version over the past years, but some just don't have the finances to do so. You best choose with your gut if you want to be on such a server.

[Wouterrinos]

22 hours ago, De Baard said:

No, my email address has dots in it as well, and it works fine, so that ain't it. What Rosetta does though is check all incoming email addresses against the stopforumspam filter which may hold your email domain. Protonmail is one of the mail addresses that cannot be used there. You also can't use a VPN when signing up.

 

But a very simple solution if you haven't tried yet, is to make an account via BOINC Manager->Add project wizard. That doesn't get the rigorous check, as far as I know.

I tried turning of VPN and it worked! Thanx. Running Rosetta now from 23-7 and also added Einstein and Collatz for later.

[yoerik]

On 3/23/2020 at 10:33 PM, Ithanul said:

There is plans to run a mini BOINC event, but that has been push off to not overlap with the current F@H event.

The plan was to run a mini event as a warm up for the upcoming Pentathlon 2020 for BOINC.

Looking forward to it - especially with Rosetta planning a challenge for Apr 4th-11th. It'll be interesting to see what happens, and how we do.

[lexxR]

On 3/28/2020 at 11:29 AM, Crabbular said:

I got one of those dodgy x79 AliExpress motherboards and an old Xeon E5-2648L for more BOINC, was going to use it as a virtualized router and pi hole etc. but that can wait.
 

I think the memory is running quad-channel, but I actually have no idea 

it likely be dual channel (has 4 slots but wired in dual channel, 2 slots per channel) i assume it's the greenish coloured motherboard with miss match chipset (huananzhi X79 motherboard) 95% sure its wired for dual channel only

 

i joined linus team for bonic (given up on folding@home until they can be bothered to sort out the client to try All servers for a work unit before giving up for retry delay or i can work out how i used to que up work units)

[Crabbular]

22 minutes ago, lexxR said:

it likely be dual channel (has 4 slots but wired in dual channel, 2 slots per channel) i assume it's the greenish coloured motherboard with miss match chipset (huananzhi X79 motherboard) 95% sure its wired for dual channel only

This one is a Jingsha mATX, it said quad-channel on the listing, and there's an option for 4x channel interleaving which I've enabled, but I don't know if it made any difference.  

Haven't booted Windows on it yet, and can't get the memory channel info under Linux through the usual methods, because the OEM information was never filled in, and the chipset doesn't match the CPU etc. there's just placeholder info there. 

So, I do think this particular board is meant to support quad channel, some of them claim that at the very least, but I just can't get the info I need under Linux. 

[Favebook]

35 minutes ago, lexxR said:

it likely be dual channel (has 4 slots but wired in dual channel, 2 slots per channel) i assume it's the greenish coloured motherboard with miss match chipset (huananzhi X79 motherboard) 95% sure its wired for dual channel only

 

i joined linus team for bonic (given up on folding@home until they can be bothered to sort out the client to try All servers for a work unit before giving up for retry delay or i can work out how i used to que up work units)

It's a little bit better situation on F@H servers, and it seems to be improving as days go by. You can jump in back to check if you want.

[LIGISTX]

Not sure if this is the place to post... But I am seemingly not getting more jobs for Rosseta. Both of my rigs I have the CPU's running Rosetta and both are not queuing up new jobs. Known issue? Too many users like FAH?

 

Also, looks like if I hit properties, I have 117 tasks completed, 11 failed. What would case a failed task? I am running an overclock that has never failed a stress test, seemingly at least. Would this be indicative of an unstable OC, or can tasks fail for other reasons?