mihaibujanca / dynamicfusion

Implementation of Newcombe et al. CVPR 2015 DynamicFusion paper
BSD 3-Clause "New" or "Revised" License
391 stars 105 forks source link

KinFu2 error: no kernel image is available for execution on the device #68

Closed jiwonmoon closed 5 years ago

jiwonmoon commented 5 years ago

I execute the dynamicfusion. but, it has an error.


$ ./build/bin/dynamicfusion data/umbrella

KinFu2 error: no kernel image is available for execution on the device /home/jiwon/jiwon/dynamicfusion/kfusion/src/cuda/tsdf_volume.cu:40

this is error message.

and i do this twice in computers the first one is gpu(mx150) and another is gpu(1080)

and in the /tsdf_volume.cu:40. the error line is


cudaSafeCall ( cudaGetLastError () );

mihaibujanca commented 5 years ago

Hi "no kernel image is available for execution on the device" usually happens when either NVIDIA drivers don't work or when the compute architecture doesn't match the cuda version.

Could you please specify your cuda version, whether or not you can run the cuda examples or any cuda code?

jiwonmoon commented 5 years ago

Hi. thank you i solve the problem

to use cuda 8.0 and opencv 2.4 version

thank you!

-----Original Message----- From: "Mihai Bujanca"notifications@github.com To: "mihaibujanca/dynamicfusion"dynamicfusion@noreply.github.com; Cc: "jiwon"mjw314@naver.com; "Author"author@noreply.github.com; Sent: 2018-11-07 (수) 23:48:58 Subject: Re: [mihaibujanca/dynamicfusion] KinFu2 error: no kernel image is available for execution on the device (#68)

Hi "no kernel image is available for execution on the device" usually happens when either NVIDIA drivers don't work or when the compute architecture doesn't match the cuda version. Could you please specify your cuda version, whether or not you can run the cuda examples or any cuda code? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

remmel commented 6 months ago

Check your NVidia card compute capability in https://en.m.wikipedia.org/wiki/CUDA#GPUs_supported and make sure that it is in https://github.com/mihaibujanca/dynamicfusion/blob/b687a8987b77c9f613ed072726729d8ce05a458a/CMakeLists.txt#L63 eg my RTX3060 is compute capability 8.6