Is the GPU application going to be updated to allow the use of the new GPU's from Nvidia like the GTX 1660, and the RTX versions ??

They currently do not work with this application.

Task http://asteroidsathome.net/boinc/result.php?resultid=243597321

---

Proud Founder of
Have a look at my WebCam

From what little I do know about running programs on a GPU, I don't understand why a new card would not be able to run an application that runs on a slightly older card. I mean, if it was a radically different architecture or different paradigm or something like that I would understand, but from what I understand Nvidia is pretty much sticking with the cuda architecture. I know that they have different generations of compute capability, but shouldn't an application that relies on older capabilities still run on a card that has all of those, plus new ones?

I don't know why, but I find the whole idea of GPGPU computing really interesting. In fact, any kind of system where the primary processor offloads some of the workload to a totally independent system is fascinating to me.

Well they still do not work. I jst got this error:

<core_client_version>7.14.2</core_client_version>
<![CDATA[
<message>
The system cannot find the file specified.
(0x2) - exit code 2 (0x2)</message>
<stderr_txt>
CUDA RC12!!!!!!!!!!
CUDA Device number: 0
CUDA Device: GeForce GTX 1660 Ti
Compute capability: 7.5
Multiprocessors: 24
Unsupported CC detected (CC2.0 and better supported only).

</stderr_txt>
]]>



http://asteroidsathome.net/boinc/results.php?userid=76

---

Proud Founder of
Have a look at my WebCam

Well they still do not work. I jst got this error:

<core_client_version>7.14.2</core_client_version>
<![CDATA[
<message>
The system cannot find the file specified.
(0x2) - exit code 2 (0x2)</message>
<stderr_txt>
CUDA RC12!!!!!!!!!!
CUDA Device number: 0
CUDA Device: GeForce GTX 1660 Ti
Compute capability: 7.5
Multiprocessors: 24
Unsupported CC detected (CC2.0 and better supported only).

</stderr_txt>
]]>



http://asteroidsathome.net/boinc/results.php?userid=76

Could it be "Creative Commons"?
https://creativecommons.org/about/

If so that probably is't your problem.

Could it be "Creative Commons"?
https://creativecommons.org/about/

If so that probably is't your problem.

Actually it stands for Compute Capability. The version of the Cuda toolkit being used does not support the card in question. The developers need to update the Cuda toolkit being used in order to support newer GPUs.

The long term solution would be to port the application to OpenCL. When Pascal was released, it only supported OpenCL initially until Nvidia released a new version of CUDA. What I have read about CUDA is that it encourages developers to target a specific hardware. A developer could target one type of Nvidia GPU hardware at the expense of others, which will either force the CUDA driver to emulate the code to fit the mismatched hardware if a mismatched GPU is found, or refuse to run it if there is no way to allow the hardware mismatch to be solved. OpenCL can allow the developer to write generic code that runs mostly equally well on many different hardware architectures. OpenCL also allows AMD GPUs to participate. Nvidia has been sluggish to adopt new versions of OpenCL, but it does not skimp on supporting OpenCL 1.2 or older. Nvidia's drivers also deprecate the running of and sometimes remove the support for running old CUDA versions on newer GPUs.

I wonder why no developer, Project Admin is making any remarks to this problem. Post 6317 was posted in Jul with a suggestion that might work yet we have no input from the Admins of this project since that time.

I wonder what we should in in the absence of an Admin. Should we boycott the project because of no Admin Support. After-all the cost of a new GPU is not cheap and if used it would facilitate a faster completion of the Task, benefiting the project. If I am to take my new GTX 1660 to a new project, I would not bother staying here just to crunch CPU.

---

Jesse,

Also, porting to OpenCL would allow AMD Radeon cards to crunch here, as well.

James

---

It's almost been a year since the release of these new cards and not a word from the administration. With that being said, a gtx 1080 is about 3 x a single thread in terms of speed at this point, so maybe they're trying to work on the entire application and make it faster?

Either way it's very disheartening not getting a single update. I finally decided to crunch here for the long term, but after going in depth and browsing these forums, I believe I'll meet my goal and find somewhere else to crunch.

Rant time: Maybe it is a lack of funding. Maybe it is a lack of time. But no communication whatsoever after we, as crunchers, devote thousands of hours (as well as in some cases dollars) into builds, electrical costs, etc. Sorry, that isn't right.

Hey Asteroid-Man, When are you going to compile your code for CUDA 10 so the Turing GPUs will work???

---

Yes, is it just me, or does Asteroid not support GTX 2080 Ti? Is this a confirmed / known issue? It looks like it may be. I'd love to know a workaround if it is!

I am receiving the below error, although the GPU works fine on other project GPU applications.

<core_client_version>7.14.2</core_client_version>
<![CDATA[
<message>
The system cannot find the file specified.
(0x2) - exit code 2 (0x2)</message>
<stderr_txt>
CUDA RC12!!!!!!!!!!
CUDA Device number: 0
CUDA Device: GeForce RTX 2080 Ti
Compute capability: 7.5
Multiprocessors: 68
Unsupported CC detected (CC2.0 and better supported only).

</stderr_txt>
]]>

This project is also erroring on my RTX 2080.

Can this be fixed, please?

http://asteroidsathome.net/boinc/result.php?resultid=267662207

<core_client_version>7.14.2</core_client_version>
<![CDATA[
<message>
The system cannot find the file specified.
(0x2) - exit code 2 (0x2)</message>
<stderr_txt>
CUDA RC12!!!!!!!!!!
CUDA Device number: 0
CUDA Device: GeForce RTX 2080
Compute capability: 7.5
Multiprocessors: 46
Unsupported CC detected (CC2.0 and better supported only).

</stderr_txt>
]]>

Admins got hit by an asteroid. No need to run this project anymore.

Cut the crap - someone along the line is not supporting the concept of -
"backwards compatibility"
I'm no expert - but the references seem to point at David Anderson at BOINC

Yea, just discovered the problem. My gtx 1660ti error'ed out 50+ tasks within seconds of attaching the project.

must have double clicked, sorry, delete one of these assume any moderators or principals care.

Yea, just discovered the problem. My gtx 1660ti error'ed out 50+ tasks within seconds of attaching the project.

I haven't had any issues running it on my RTX 2060 (4GB) as far as I can tell.