As of today AMD CPUs are simply the best for distributed computing, no matter what, be it Ryzen, Threadripper or EPYC. Their total computing power is simply unmatched.

---

As of today AMD CPUs are simply the best for distributed computing, no matter what, be it Ryzen, Threadripper or EPYC. Their total computing power is simply unmatched.

If it is CPU only.
For GPUs, AMD are best if Double Precision is required (unless you want to forkout for workstation hardware). For Single Precision, Nvidia.

---

Grant
Darwin NT

As of today AMD CPUs are simply the best for distributed computing, no matter what, be it Ryzen, Threadripper or EPYC. Their total computing power is simply unmatched.

If it is CPU only.
For GPUs, AMD are best if Double Precision is required (unless you want to forkout for workstation hardware). For Single Precision, Nvidia.

It also depends on if you can run all the date being processed in the gpu in question. If not then "he" is right. If I understand the way Rosetti is approaching the processing using larger databases instead of discrete data packages then that design for data-analysis will not show the true benefits of the massive parallel processing offered by gpus.

Tom M

---

Help, my tagline is missing..... Help, my tagline is......... Help, m........ Hel.....

As of today AMD CPUs are simply the best for distributed computing, no matter what, be it Ryzen, Threadripper or EPYC. Their total computing power is simply unmatched.

If it is CPU only.
For GPUs, AMD are best if Double Precision is required (unless you want to forkout for workstation hardware). For Single Precision, Nvidia.

It also depends on if you can run all the date being processed in the gpu in question. If not then "he" is right. If I understand the way Rosetti is approaching the processing using larger databases instead of discrete data packages then that design for data-analysis will not show the true benefits of the massive parallel processing offered by gpus.

Tom M

Sorry, "data"

---

Help, my tagline is missing..... Help, my tagline is......... Help, m........ Hel.....

Of course my comment is only about choosing between Intel and AMD CPUs for CPU only workloads.

---

Of course my comment is only about choosing between Intel and AMD CPUs for CPU only workloads.

Point.
For the average individual user who probably has a several year old Intel cpu "you dance with who brung ya" For all the new fangled stuff as long as we aren't struggling with some version of the AVX instructions the evidence reported in the reviews does support it.

Tom M

---

Help, my tagline is missing..... Help, my tagline is......... Help, m........ Hel.....

Greetings,

I got it under "control". I figured out what to do to get beyond the issues I was having.

1. ) Set Get New Tasks.
2. ) Once downloaded, set NNT.
3. ) If tasks are getting really close to the finish line, DON'T go into Windows 10 to do anything in WoW. Let them finish and then go into Windows 10.
4. ) Once cache is depleted, set get new tasks, download and then reset NNT.

Setting no new tasks allows the tasks to run in normal mode and not high priority since no tasks are added to cache with a close deadline. Tasks runtime is now estimated just under 4 hours. Sure beats the 8 hours before. ;)

Since I'm only setup to get GPU tasks from SETI, I won't have to worry about Rosetta tasks going into high priority, plus I'm only using 6 cores on Rosetta which leaves CPU cores to cover the GPU SETI tasks.

I told you I really wanted to do this project; I just needed to figure out a plan of attack to do it.

Have a great day! :)

Siran

---

CAPT Siran d'Vel'nahr XO
USS Vre'kasht NCC-33187

"Logic is the cement of our civilization with which we ascend from chaos using reason as our guide." - T'Plana-hath

I got it under "control". I figured out what to do to get beyond the issues I was having.

From another thread on the no Linux checkpoint issue.

<cc_config>
    <options>
        <no_alt_platform>1</no_alt_platform>
    </options>
</cc_config>

in app_config.xml apparently stops the Manager from using the Application that has the checkpointing issue.

Tasks runtime is now estimated just under 4 hours. Sure beats the 8 hours before. ;)

So you changed the Target CPU Runtime to 4 hours? If not, they will still run for whatever it is set for (8 hrs is the default).

---

Grant
Darwin NT

I got it under "control". I figured out what to do to get beyond the issues I was having.

From another thread on the no Linux checkpoint issue.

<cc_config>
    <options>
        <no_alt_platform>1</no_alt_platform>
    </options>
</cc_config>

in app_config.xml apparently stops the Manager from using the Application that has the checkpointing issue.

Tasks runtime is now estimated just under 4 hours. Sure beats the 8 hours before. ;)

So you changed the Target CPU Runtime to 4 hours? If not, they will still run for whatever it is set for (8 hrs is the default).

Hi Grant,

Yes, I change the CPU runtime to 4 hours yesterday. The first batch of tasks I got were estimated just under 8 hours and the batch this morning just under 4.

This is my cc_config.xml currently:

<cc_config>
 <log_flags>
   <sched_op_debug>1</sched_op_debug>
 </log_flags>
 <options>
   <use_all_gpus>1</use_all_gpus>
   <exclude_gpu>
    <url>http://setiathome.berkeley.edu/</url>
    <device_num>0</device_num>
    <type>NVIDIA</type>
   </exclude_gpu>
   <exclude_gpu>
    <url>http://einstein.phys.uwm.edu/</url>
    <device_num>0</device_num>
    <type>NVIDIA</type>
   </exclude_gpu>
   <save_stats_days>365</save_stats_days>
   <allow_remote_gui_rpc>1</allow_remote_gui_rpc>
 </options>
</cc_config>

If I add the line you mentioned, will it affect SETI or even Einstein (which I'm not running)?

[edit]
Wait a minute. You listed <cc_config> ... </cc_config>, but you mentioned app_config.xml?
[/edit]
[edit2]
Ok, I found the posts talking about that option and added it to my cc_config.xml file and reloaded.
[/edit2]

Have a great day! :)

Siran

---

CAPT Siran d'Vel'nahr XO
USS Vre'kasht NCC-33187

"Logic is the cement of our civilization with which we ascend from chaos using reason as our guide." - T'Plana-hath

Just looked at the tasks I am crunching this morning.
Even though I was set to the "default" (8 hours?) the majority of the tasks are now running near 12 hours.

Is this a bubble or a trend?

Tom

---

Help, my tagline is missing..... Help, my tagline is......... Help, m........ Hel.....

The few WUs on this host that I see that ran for 12 hours, using the Rosetta v4.15
x86_64-pc-linux-gnu application, and were all ended by the watchdog. I see the host has 32 CPUs and 32GB of memory. How many active R@h threads are running at the same time on this host?

---

Rosetta Moderator: Mod.Sense

The few WUs on this host that I see that ran for 12 hours, using the Rosetta v4.15
x86_64-pc-linux-gnu application, and were all ended by the watchdog. I see the host has 32 CPUs and 32GB of memory. How many active R@h threads are running at the same time on this host?

18 threads with no pauses.

I have examined the tasks and they all seem to be "i686" tasks. I thought I had disabled those with the no_alt_platform> but apparently not.
---edit----------
Based on suggestion(s) from other threads I have aborted the "i686" tasks which were all 4.15 versions.
Now what? I have gone to NNT while I see what else I can try
---edit---

Tom M

---

Help, my tagline is missing..... Help, my tagline is......... Help, m........ Hel.....

Yes, I stand corrected. I was looking at the "Application version" section in the header, not the command string in the output section. The command string shows i686.

I am not certain why the cc_config doesn't seem to do what folks told me it would do.

---

Rosetta Moderator: Mod.Sense

Yes, I stand corrected. I was looking at the "Application version" section in the header, not the command string in the output section. The command string shows i686.

I am not certain why the cc_config doesn't seem to do what folks told me it would do.

After confirming I had that parameter in there I cycled the Boinc Manager/Clients to make sure the cc_config.xml file was read.
I am going to throttle my # of threads to 8 and my time to 2 hours and then allow new tasks again. I will take a look at the stuff after it downloads and starts running and see if it avoid the i686.

My impression was the "mini" shows up more often when the short tasks are selected so I may be able to screen it out that way. Or at least reduce the time before the watch dog timer catches up with it.

Tom M

---

Help, my tagline is missing..... Help, my tagline is......... Help, m........ Hel.....

Tom, if it doesn't work, I would empty the cache and try resetting the project, might work.

Tom, if it doesn't work, I would empty the cache and try resetting the project, might work.

Just did that and it appears to be only downloading non-"686" tasks now for both mini and regular rosetti@home.

Thank you for the idea.
I am still running "only" 8 threads while I run down the tasks on another cpu & gpu project that I have switched to gpu only.

OBTW, what was the "advantage" of the "686" version over the version I am running now?

Tom M

---

Help, my tagline is missing..... Help, my tagline is......... Help, m........ Hel.....

OBTW, what was the "advantage" of the "686" version over the version I am running now?

None.
One is 64bit, one is 32bit, but the performance of each application is pretty much the same as the other.

---

Grant
Darwin NT

Greetings,

I just discovered something that seems a bit too weird to me.

My task list here says the tasks I got this morning are the x86_64 64 bit tasks. Yet, my tasks properties say they are i686 tasks. Why is that? Or is it that the tasks are intelligent and decide which app is gonna crunch them? lol ;) Or is it that the server itself does not know which app is gonna crunch the tasks so it shows the better of the 2? lol ;)

Something else I don't quite understand is why, if I'm running a 64 bit processor and a 64 bit OS (Linux Mint), do I run a 32 bit app for these tasks?

Not really complaining since I'm happy to be crunching here. Just curious. :)

Have a great day! :)

Siran

---

CAPT Siran d'Vel'nahr XO
USS Vre'kasht NCC-33187

"Logic is the cement of our civilization with which we ascend from chaos using reason as our guide." - T'Plana-hath

Greetings,

I just discovered something that seems a bit too weird to me.

My task list here says the tasks I got this morning are the x86_64 64 bit tasks. Yet, my tasks properties say they are i686 tasks. Why is that? Or is it that the tasks are intelligent and decide which app is gonna crunch them? lol ;) Or is it that the server itself does not know which app is gonna crunch the tasks so it shows the better of the 2? lol ;)

Something else I don't quite understand is why, if I'm running a 64 bit processor and a 64 bit OS (Linux Mint), do I run a 32 bit app for these tasks?

Not really complaining since I'm happy to be crunching here. Just curious. :)

Have a great day! :)

Siran

Same thing happened to me after I put the no_alt_platform in cc_config, BOINC deleted the 86_64 app and used the i686 app for all tasks instead. I see you have some tasks only paying 20 credits and running much longer than they should, so I guess you have the same problem.

Solution I think, is either delete all tasks or select NNT and empty cache, then do a project reset and see if it works...

Greetings,

I just discovered something that seems a bit too weird to me.

My task list here says the tasks I got this morning are the x86_64 64 bit tasks. Yet, my tasks properties say they are i686 tasks. Why is that? Or is it that the tasks are intelligent and decide which app is gonna crunch them? lol ;) Or is it that the server itself does not know which app is gonna crunch the tasks so it shows the better of the 2? lol ;)

Something else I don't quite understand is why, if I'm running a 64 bit processor and a 64 bit OS (Linux Mint), do I run a 32 bit app for these tasks?

Not really complaining since I'm happy to be crunching here. Just curious. :)

Have a great day! :)

Siran

Same thing happened to me after I put the no_alt_platform in cc_config, BOINC deleted the 86_64 app and used the i686 app for all tasks instead. I see you have some tasks only paying 20 credits and running much longer than they should, so I guess you have the same problem.

Solution I think, is either delete all tasks or select NNT and empty cache, then do a project reset and see if it works...

Hi John,

Yep, I added the no_alt_platform to my cc_config.xml file. I too saw the 20 credit tasks. I shall do as you suggested, if'n I remember to. ;) I'd rather just empty my cache and then reset instead of aborting the tasks I have. I have 27 of them.

Have a great day! :)

Siran

---

CAPT Siran d'Vel'nahr XO
USS Vre'kasht NCC-33187

"Logic is the cement of our civilization with which we ascend from chaos using reason as our guide." - T'Plana-hath