I would say there is an improvement in the 0.98 code compared to 0.97. I proved that to myself when I benched the betas for Petri before wide release. But Ian states that his beta of the 0.98b1 CUDA10 with driver version 410 beats out the 0.98b1 CUDA101 version with drivers 418. He kept his beta release and is running that to great effect. I was not a tester of that version so never got a chance to benchmark it.

. . Would you say that improvement (0.98b over 0.97) is enough to bother updating? I am running 0.97 on both the rig with 970s and the rig with the 1050ti (which sadly is using the repository version of BOINC) so after reading things like the comment that the 101 version is only useful on Turing cards I am not confident it would make any significant difference on my rigs.

Stephen

? ?

it's absolutely worth updating. just because there is still a CUDA 9.0 version of the app, doesn't mean it's the same app. v0.98 is quite a bit faster than v0.97.

if you have a Turing card, then the CUDA 10+ apps are very slightly faster than 9.0. but v0.98 is faster than 0.97 all else being equal.

---

Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

I would say there is an improvement in the 0.98 code compared to 0.97. I proved that to myself when I benched the betas for Petri before wide release. But Ian states that his beta of the 0.98b1 CUDA10 with driver version 410 beats out the 0.98b1 CUDA101 version with drivers 418. He kept his beta release and is running that to great effect. I was not a tester of that version so never got a chance to benchmark it.

. . Would you say that improvement (0.98b over 0.97) is enough to bother updating? I am running 0.97 on both the rig with 970s and the rig with the 1050ti (which sadly is using the repository version of BOINC) so after reading things like the comment that the 101 version is only useful on Turing cards I am not confident it would make any significant difference on my rigs.

Stephen

? ?

it's absolutely worth updating. just because there is still a CUDA 9.0 version of the app, doesn't mean it's the same app. v0.98 is quite a bit faster than v0.97.

if you have a Turing card, then the CUDA 10+ apps are very slightly faster than 9.0. but v0.98 is faster than 0.97 all else being equal.

So for a Linux Tbar system that is running strickly gtx type nvidia video cards then going back to the 98? version included in the setup 10 and downgrading to one step below the 4.18 video driver is a good idea? Or did I miss something?

---

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

I'm not sure I understand what you're asking.

V0.98b1 CUDA 9.0 is faster than V0.97 CUDA 9.0
V0.98b1 CUDA 10.0 is faster than V0.97 CUDA 10.0
CUDA 10 (or 10.1) vs CUDA 9 is probably negligible, only showing a slight advantage to the Turing cards
the v0.98 app is anywhere from 10-30% faster on average than v0.97 in my tests (varied by WU and type)


I found the best performance with the CUDA 10 app over the 10.1 app, but the difference has little or nothing to do with the app itself. The difference is due to the drivers needed.

I found upwards of 5% reduction in processing speed when using drivers 415+. Nvidia did "something" that caused this slowdown. since the CUDA 10.1 app requires 418+ drivers to be used, you get hit with this slowdown from the DRIVERS, not the app. So since all of my cards now are early versions of Turing that can use the 410 drivers, my personal preferred setup is the CUDA 10 app with 410 drivers. If you have any of the later Turing cards (RTX 2060, GTX 16-series, RTX Super series) you will be required to use the later drivers for compatibility anyway.

the v0.98 app is anywhere from 10-30% faster on average than v0.97 in my tests (varied by WU and type)

if you have a GTX 10-series card, I would probably use the v0.98b1 CUDA 9.0 app with drivers 410 or lower.

---

Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

I'm not sure I understand what you're asking.

You nailed the answer though. Thank you.
I will see if I can migrate back down to 4.15

Tom

---

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

no, you need 410 or lower. 415 and later is when i noticed the slowdown. also make sure you're pairing that with the CUDA 90 app, as the CUDA 10.1 app wont run on any driver less than 418.

---

Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

no, you need 410 or lower. 415 and later is when i noticed the slowdown. also make sure you're pairing that with the CUDA 90 app, as the CUDA 10.1 app wont run on any driver less than 418.

Thank you for the correction.

As far as I can tell, the CUDA10 archive has a "cuda9" in it that will run 4.10?

Tom

---

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

Well I finally upgraded the drivers and now I am crunching, for better or worse, with CUDA101 on all my hosts that have the capability to do so.

i don't know what you mean by "CUDA10 archive"

but the app you want to use with the 410 (or lower) drivers will be the CUDA90 app. This is CUDA 9.0 enabled. not CUDA 10. I have a beta version that is CUDA 10 (not 10.1) enabled.

the apps available to you in the latest standard All-In-One package are the CUDA 9.0 and CUDA 10.1 app. choose the 9.0 one. You wont see any benefit using a CUDA 10+ app on your GTX 10-series cards.

---

Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

So any idea how much of a gain there is going from cuda10.1 to cuda9 on a 10x0 only system?

probably none by just changing the app alone.

but you could get some extra speed by changing the app AND the drivers down to 410 or lower (and then using the 9.0 app, since the 10.1 app wont work with the older drivers). the speed boost will come from the drivers, not the app.

---

Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

i don't know what you mean by "CUDA10 archive"

but the app you want to use with the 410 (or lower) drivers will be the CUDA90 app. This is CUDA 9.0 enabled. not CUDA 10. I have a beta version that is CUDA 10 (not 10.1) enabled.

the apps available to you in the latest standard All-In-One package are the CUDA 9.0 and CUDA 10.1 app. choose the 9.0 one. You wont see any benefit using a CUDA 10+ app on your GTX 10-series cards.

I downloaded a new copy. And it appears I was running an unknown previous version of the 10.1 archive.
It was throwing computation errors.

Installed the "new" archive over the top of the "old" stuff. Put in the "-nobs" in the app_info.xml file.

Running it now with NNT while I confirm it has stopped throwing "computation errors".

Tom

---

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

My new host wont see the GPU in ubuntu. What am I did wrong?

quotro P2000

event says app version refers to missing nvidia gpu

---

My new host wont see the GPU in ubuntu. What am I did wrong?

Did you install the Nvidia drivers? If you did. . . . what does nvidia-smi show? What does Nvidia X Server Settings show? Have you installed clinfo to allow easy checking of installed CUDA and OpenCL drivers?
The Quadro K2000 is not able to use the special app since the card only has CC = 3.0. You will have to use the stock SoG application which requires OpenCL.

---

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

A proud member of the OFA (Old Farts Association)

My new host wont see the GPU in ubuntu. What am I did wrong?

Did you install the Nvidia drivers? If you did. . . . what does nvidia-smi show? What does Nvidia X Server Settings show? Have you installed clinfo to allow easy checking of installed CUDA and OpenCL drivers?
The Quadro K2000 is not able to use the special app since the card only has CC = 3.0. You will have to use the stock SoG application which requires OpenCL.

Drivers are UP, sorry its a P2000 not K2000.

---

My new host wont see the GPU in ubuntu. What am I did wrong?

Did you install the Nvidia drivers? If you did. . . . what does nvidia-smi show? What does Nvidia X Server Settings show? Have you installed clinfo to allow easy checking of installed CUDA and OpenCL drivers?
The Quadro K2000 is not able to use the special app since the card only has CC = 3.0. You will have to use the stock SoG application which requires OpenCL.

Drivers are UP, sorry its a P2000 not K2000.

OK, sorry about the confusion. Able to use the special app. What does any of the programs I mentioned show for drivers? Which Nvidia version did you install?

Have you rebooted the system yet after installing the Nvidia drivers?

---

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

A proud member of the OFA (Old Farts Association)

My new host wont see the GPU in ubuntu. What am I did wrong?

Did you install the Nvidia drivers? If you did. . . . what does nvidia-smi show? What does Nvidia X Server Settings show? Have you installed clinfo to allow easy checking of installed CUDA and OpenCL drivers?
The Quadro K2000 is not able to use the special app since the card only has CC = 3.0. You will have to use the stock SoG application which requires OpenCL.

Drivers are UP, sorry its a P2000 not K2000.

OK, sorry about the confusion. Able to use the special app. What does any of the programs I mentioned show for drivers? Which Nvidia version did you install?

nvidia X server is also showing the gpu as P2000.
Have you rebooted the system yet after installing the Nvidia drivers?

precision@precision-Precision-3630-Tower:~$ lspci -vnn | grep -i VGA -A 12
01:00.0 VGA compatible controller [0300]: NVIDIA Corporation GP106GL [Quadro P2000] [10de:1c30] (rev a1) (prog-if 00 [VGA controller])
Subsystem: Dell GP106GL [Quadro P2000] [1028:11b3]
Flags: bus master, fast devsel, latency 0, IRQ 141
Memory at a3000000 (32-bit, non-prefetchable) [size=16M]
Memory at 90000000 (64-bit, prefetchable) [size=256M]
Memory at a0000000 (64-bit, prefetchable) [size=32M]
I/O ports at 3000 [size=128]
Expansion ROM at 000c0000 [disabled] [size=128K]
Capabilities: <access denied>
Kernel driver in use: nouveau
Kernel modules: nvidiafb, nouveau, nvidia_drm, nvidia

01:00.1 Audio device [0403]: NVIDIA Corporation GP106 High Definition Audio Controller [10de:10f1] (rev a1)

Driver 418.

Seems GPU was picked up, after bios update

NVIDIA Quadro P2000 (4095MB) driver: 418.56 OpenCL: 1.2

---

probably none by just changing the app alone.

but you could get some extra speed by changing the app AND the drivers down to 410 or lower (and then using the 9.0 app, since the 10.1 app wont work with the older drivers). the speed boost will come from the drivers, not the app.

Yes of course with changing drivers. I'm asking because if the gain is very small I'm not going to change anything.

My new host wont see the GPU in ubuntu. What am I did wrong?

Did you install the Nvidia drivers? If you did. . . . what does nvidia-smi show? What does Nvidia X Server Settings show? Have you installed clinfo to allow easy checking of installed CUDA and OpenCL drivers?
The Quadro K2000 is not able to use the special app since the card only has CC = 3.0. You will have to use the stock SoG application which requires OpenCL.

Drivers are UP, sorry its a P2000 not K2000.

OK, sorry about the confusion. Able to use the special app. What does any of the programs I mentioned show for drivers? Which Nvidia version did you install?

Have you rebooted the system yet after installing the Nvidia drivers?

precision@precision-Precision-3630-Tower:~$ lspci -vnn | grep -i VGA -A 12
01:00.0 VGA compatible controller [0300]: NVIDIA Corporation GP106GL [Quadro P2000] [10de:1c30] (rev a1) (prog-if 00 [VGA controller])
Subsystem: Dell GP106GL [Quadro P2000] [1028:11b3]
Flags: bus master, fast devsel, latency 0, IRQ 141
Memory at a3000000 (32-bit, non-prefetchable) [size=16M]
Memory at 90000000 (64-bit, prefetchable) [size=256M]
Memory at a0000000 (64-bit, prefetchable) [size=32M]
I/O ports at 3000 [size=128]
Expansion ROM at 000c0000 [disabled] [size=128K]
Capabilities: <access denied>
Kernel driver in use: nouveau
Kernel modules: nvidiafb, nouveau, nvidia_drm, nvidia

01:00.1 Audio device [0403]: NVIDIA Corporation GP106 High Definition Audio Controller [10de:10f1] (rev a1)

Driver 418.

Look at what I highlighted RED

You have not rebooted the system since installing the 418 drivers. You have to reboot the system to use or change over to any different driver since the drivers are incorporated into a new kernel that is built when you change drivers.

---

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

A proud member of the OFA (Old Farts Association)

My new host wont see the GPU in ubuntu. What am I did wrong?

Did you install the Nvidia drivers? If you did. . . . what does nvidia-smi show? What does Nvidia X Server Settings show? Have you installed clinfo to allow easy checking of installed CUDA and OpenCL drivers?
The Quadro K2000 is not able to use the special app since the card only has CC = 3.0. You will have to use the stock SoG application which requires OpenCL.

Drivers are UP, sorry its a P2000 not K2000.

OK, sorry about the confusion. Able to use the special app. What does any of the programs I mentioned show for drivers? Which Nvidia version did you install?

Have you rebooted the system yet after installing the Nvidia drivers?

precision@precision-Precision-3630-Tower:~$ lspci -vnn | grep -i VGA -A 12
01:00.0 VGA compatible controller [0300]: NVIDIA Corporation GP106GL [Quadro P2000] [10de:1c30] (rev a1) (prog-if 00 [VGA controller])
Subsystem: Dell GP106GL [Quadro P2000] [1028:11b3]
Flags: bus master, fast devsel, latency 0, IRQ 141
Memory at a3000000 (32-bit, non-prefetchable) [size=16M]
Memory at 90000000 (64-bit, prefetchable) [size=256M]
Memory at a0000000 (64-bit, prefetchable) [size=32M]
I/O ports at 3000 [size=128]
Expansion ROM at 000c0000 [disabled] [size=128K]
Capabilities: <access denied>
Kernel driver in use: nouveau
Kernel modules: nvidiafb, nouveau, nvidia_drm, nvidia

01:00.1 Audio device [0403]: NVIDIA Corporation GP106 High Definition Audio Controller [10de:10f1] (rev a1)

Driver 418.

Look at what I highlighted RED

You have not rebooted the system since installing the 418 drivers. You have to reboot the system to use or change over to any different driver since the drivers are incorporated into a new kernel that is built when you change drivers.

Lesson learned, thank you

---

On My machines I see around a 10% speedup on the Arecibo tasks, plus you can now run a GTX 960 2 GB without it running out of vRAM at unroll 8. Those items alone should be enough for everyone to update. If you have a Turing GPU the BLC tasks also see a nice speed improvement whereas it's not quite so much on the older GPUs. My 750 Ti sees a BLC speedup, the 970 doesn't see a BLC speedup, EVERYTHING sees an Arecibo Speedup.

. . Thanks for that TBar, I take it that comparison is between 0.98b1-90 and 0.98b1-101. I do remember the mention of the lower rate of inconclusives but with very little Arecibo work and an otherwise low rate already it was not compelling for me. I guess it is doable then on both my machines with Ubuntu 14.04. Again thanks.

. . One thing that I find amusing is that there was an improvement on GBT WUs on the 750ti but not the 970s, those little cards are amazing. Such low power usage but very impressive performance.

Stephen

:)