The 2990W I have coming online "soon" is reusing a dual-boot Win10 and Lubuntu HD. I assume I would need an upgrade to the Lunatics distro to test your newest crunching app?

No, this app needs to get installed manually.

---

With each crime and every kindness we birth our future.

I would like to see how the Threadripper performs with the new app i made available on my website.
Sadly the 2990WX runs on Linux.

What app is that? I don't see anything new at your website.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

I would like to see how the Threadripper performs with the new app i made available on my website.
Sadly the 2990WX runs on Linux.

What app is that? I don't see anything new at your website.

Its r3714_Vlock.
Maybe you need to refresh the page after loading.
Its been downloaded a few times already.

---

With each crime and every kindness we birth our future.

The 2990W I have coming online "soon" is reusing a dual-boot Win10 and Lubuntu HD. I assume I would need an upgrade to the Lunatics distro to test your newest crunching app?

No, this app needs to get installed manually.

I think I understand. Install the Lunatics Distro and then manually install the new app?

Tom

---

A proud member of the OFA (Old Farts Association).

Guess I missed it the 3 or 4 times I went to view the Download page. Pretty sure I didn't have an old version of the website in the browser cache since I cleared it a couple of days ago and haven't visited your site in months. See it now.

Did you write the app or Raistmer? Haven't seen any discussion by Raistmer yet in his threads.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Guess I missed it the 3 or 4 times I went to view the Download page. Pretty sure I didn't have an old version of the website in the browser cache since I cleared it a couple of days ago and haven't visited your site in months. See it now.

Did you write the app or Raistmer? Haven't seen any discussion by Raistmer yet in his threads.

It`s Raistmers app off course.
I just thought after 9 month of testing its time to try a open beta.
It should be faster on any Zen based CPU.
Not sure about Intels since i have no clue about their memory handling.

---

With each crime and every kindness we birth our future.

I assume then Raistmer offered the new beta app to you for testing. Did he tell you anything about it? Like what is V_lock doing in the name? Something to do with gpu memory from your comment I assume.

[Edit] Just looked at your hosts and noticed the cpu tasks run with the r3714 cpu app. So a cpu app and not a gpu app.

What are the clocks on the cpus? The speed isn't printed out in the stderr.txt like for the SSE41 apps.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

What is the command line you are using with the r3714 app?

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

I assume then Raistmer offered the new beta app to you for testing. Did he tell you anything about it? Like what is V_lock doing in the name? Something to do with gpu memory from your comment I assume.

[Edit] Just looked at your hosts and noticed the cpu tasks run with the r3714 cpu app. So a cpu app and not a gpu app.

What are the clocks on the cpus? The speed isn't printed out in the stderr.txt like for the SSE41 apps.

Like i said already i`m testing this app since february in closed alpha on Lunatics.
Since July i run it on beta and a few weeks later on main.
Its on me to decide if its worth to make it public.
Raistmer trusts me on this.

I`m running my Ryzen on stock speed 3,6 GHZ.

---

With each crime and every kindness we birth our future.

I assume then Raistmer offered the new beta app to you for testing. Did he tell you anything about it? Like what is V_lock doing in the name? Something to do with gpu memory from your comment I assume.

[Edit] Just looked at your hosts and noticed the cpu tasks run with the r3714 cpu app. So a cpu app and not a gpu app.

What are the clocks on the cpus? The speed isn't printed out in the stderr.txt like for the SSE41 apps.

Like i said already i`m testing this app since february in closed alpha on Lunatics.
Since July i run it on beta and a few weeks later on main.
Its on me to decide if its worth to make it public.
Raistmer trusts me on this.

I`m running my Ryzen on stock speed 3,6 GHZ.

Ok, was just trying to attribute crunching time to either the app or the cpu clocks. OK. You are running 90% as fast clocks as I am but your improvement in crunch time does not scale concurrently to the difference between the r3711 SSE41 app and your new r3714 app. So must be some other parameters that are not the same between your hosts and my hosts. Likely the memory clocks.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

I assume then Raistmer offered the new beta app to you for testing. Did he tell you anything about it? Like what is V_lock doing in the name? Something to do with gpu memory from your comment I assume.

[Edit] Just looked at your hosts and noticed the cpu tasks run with the r3714 cpu app. So a cpu app and not a gpu app.

What are the clocks on the cpus? The speed isn't printed out in the stderr.txt like for the SSE41 apps.

Like i said already i`m testing this app since february in closed alpha on Lunatics.
Since July i run it on beta and a few weeks later on main.
Its on me to decide if its worth to make it public.
Raistmer trusts me on this.

I`m running my Ryzen on stock speed 3,6 GHZ.

Ok, was just trying to attribute crunching time to either the app or the cpu clocks. OK. You are running 90% as fast clocks as I am but your improvement in crunch time does not scale concurrently to the difference between the r3711 SSE41 app and your new r3714 app. So must be some other parameters that are not the same between your hosts and my hosts. Likely the memory clocks.

I`m running 16GB 3200 CL14.
Since r3714 uses new memory handling techniques yes, the RAM makes the difference.

---

With each crime and every kindness we birth our future.

I assume then Raistmer offered the new beta app to you for testing. Did he tell you anything about it? Like what is V_lock doing in the name? Something to do with gpu memory from your comment I assume.

[Edit] Just looked at your hosts and noticed the cpu tasks run with the r3714 cpu app. So a cpu app and not a gpu app.

What are the clocks on the cpus? The speed isn't printed out in the stderr.txt like for the SSE41 apps.

Like i said already i`m testing this app since february in closed alpha on Lunatics.
Since July i run it on beta and a few weeks later on main.
Its on me to decide if its worth to make it public.
Raistmer trusts me on this.

I`m running my Ryzen on stock speed 3,6 GHZ.

Ok, was just trying to attribute crunching time to either the app or the cpu clocks. OK. You are running 90% as fast clocks as I am but your improvement in crunch time does not scale concurrently to the difference between the r3711 SSE41 app and your new r3714 app. So must be some other parameters that are not the same between your hosts and my hosts. Likely the memory clocks.

I`m running 16GB 3200 CL14.
Since r3714 uses new memory handling techniques yes, the RAM makes the difference.

Hi Mike,

I just benchmarked it on my WIN10 1950x machine that runs at 3.6GHz. It has 128GB of 2400 speed ECC udimm.

WU : blc04_2bit_blc04_guppi_57898_17662_DIAG_KIC8462852_OFF_0020.12914.818.18.27.159.vlar.WU
MB8_win_x64_AVX_VS2010_r3330.exe -verb -nog :
  Elapsed 2638.336 secs
      CPU 2633.938 secs
MB8_win_x64_AVX_VS2017_r3714_Vlock.exe   :
  Elapsed 2472.117 secs, speedup: 6.30%  ratio: 1.07x
      CPU 2467.734 secs, speedup: 6.31%  ratio: 1.07x

It looks good. I will release it on my WIN10 Machine. Do you know if Raistmer updated the optimized branch of the source code? I would like to try to compile on Linux again. I have been successful in compiling it on 18.04, but the executable finds no signal. If the optimized code is updated, I would like to try it again.

---

GitHub: Ricks-Lab
Instagram: ricks_labs

I assume then Raistmer offered the new beta app to you for testing. Did he tell you anything about it? Like what is V_lock doing in the name? Something to do with gpu memory from your comment I assume.

[Edit] Just looked at your hosts and noticed the cpu tasks run with the r3714 cpu app. So a cpu app and not a gpu app.

What are the clocks on the cpus? The speed isn't printed out in the stderr.txt like for the SSE41 apps.

Like i said already i`m testing this app since february in closed alpha on Lunatics.
Since July i run it on beta and a few weeks later on main.
Its on me to decide if its worth to make it public.
Raistmer trusts me on this.

I`m running my Ryzen on stock speed 3,6 GHZ.

Ok, was just trying to attribute crunching time to either the app or the cpu clocks. OK. You are running 90% as fast clocks as I am but your improvement in crunch time does not scale concurrently to the difference between the r3711 SSE41 app and your new r3714 app. So must be some other parameters that are not the same between your hosts and my hosts. Likely the memory clocks.

I`m running 16GB 3200 CL14.
Since r3714 uses new memory handling techniques yes, the RAM makes the difference.

I run 3466 CL14 for the system memory clocks. I figured it had something to do with memory since your stderr.txt has memory command line parameters listed.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

My cooler has arrived so there is a good chance I will get the whole thing (2990WX) assembled and running tomorrow afternoon. So I won't get to see how it actually works till after maintenance.

This system will have 2 GTX 1060 3GB's that were on a single CPU e5-2670 system I am running.

As I understand it, I will be starting with completely standard settings and let any overclocking take place at the discretion of the CPU.

Since I am reusing another systems HD, the history of tasks will be higher than 0 but the recent task level will basically start at the bottom and work its way up. I haven't investigated the "AMD" folder so I am not sure what else I will need to change to make it run. I also think this was still on the CUDA90 task so will have to upgrade it once it is running.

Tom

---

A proud member of the OFA (Old Farts Association).

The "for AMD cpus" folder just contains the r3711 cpu executable that TBar compiled with flags optimized for AMD cpus. The other one in the main directory is compiled for general Intel cpus.

I would update to the 0.97b2 CUDA92 application as long as you are running version 396 drivers or greater.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

I just benchmarked it on my WIN10 1950x machine that runs at 3.6GHz. It has 128GB of 2400 speed ECC udimm.

WU : blc04_2bit_blc04_guppi_57898_17662_DIAG_KIC8462852_OFF_0020.12914.818.18.27.159.vlar.WU
MB8_win_x64_AVX_VS2010_r3330.exe -verb -nog :
  Elapsed 2638.336 secs
      CPU 2633.938 secs
MB8_win_x64_AVX_VS2017_r3714_Vlock.exe   :
  Elapsed 2472.117 secs, speedup: 6.30%  ratio: 1.07x
      CPU 2467.734 secs, speedup: 6.31%  ratio: 1.07x

It looks good. I will release it on my WIN10 Machine. Do you know if Raistmer updated the optimized branch of the source code? I would like to try to compile on Linux again. I have been successful in compiling it on 18.04, but the executable finds no signal. If the optimized code is updated, I would like to try it again.

It's Windows-targeted update.
That's, to be useful for Linux, Linux-oriented API calls should be made instead.
App uses locked memory to circumvent possible suboptimal Windows choices for paging. Same as opt AstroPulse does.

Also I attempted to use LargePages (in another build) to save TLB cache entries and improve memory cache hits, but it seems too much efforts from user are needed to run it. It requires special user right and default UAC admin can't use such right. That is, one need to create non-administrator account and directly assing corresponding right to it.
Another way is to hack UAC management. In any case, user's action required.
Vlock can be used under usual user rights.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

Thanks for the technical description of the new app Raistmer. Good to see continued development of the science apps by our volunteers.
No real way to test how it will perform against the old r3711 Linux apps until someone is able to compile it for Linux. Wouldn't be a fair test to run under a VM I suppose.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

The "for AMD cpus" folder just contains the r3711 cpu executable that TBar compiled with flags optimized for AMD cpus. The other one in the main directory is compiled for general Intel cpus.

I would update to the 0.97b2 CUDA92 application as long as you are running version 396 drivers or greater.

I didn't realize that TBar was successful in compiling the optimized CPU app with the new GCC that supports Zen optimizations. I guess that happened while I was away from the forums for the summer. Can you provide a link to the executable you mentioned above?

---

GitHub: Ricks-Lab
Instagram: ricks_labs

I'm confused. The question was about the AMD folder in the TBar package. The AMD executables are in that folder.

The TBar packages are at CA as usual. The link is for the BOINC-All-In-One

which is the 7.8.3 version in the main directory and has the 7.4.44 version in the /Docs sub-directory. So both versions are in the package along with the AMD compiled 8.22 MB r3711 version.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

I'm confused. The question was about the AMD folder in the TBar package. The AMD executables are in that folder.

The TBar packages are at CA as usual. The link is for the BOINC-All-In-One

which is the 7.8.3 version in the main directory and has the 7.4.44 version in the /Docs sub-directory. So both versions are in the package along with the AMD compiled 8.22 MB r3711 version.

I read AMD optimized in a Ryzen thread and assumed there was a breakthrough in compiling with a newer compiler to produce Zen optimized apps. I guess I read the post with an optimistic bias. I hope to give this a try again when I get some time.

---

GitHub: Ricks-Lab
Instagram: ricks_labs