Cuda error device unavailable

Quite often, I get the CUDA library to completely fail and return with an error 46 («all CUDA-capable devices are busy or unavailable») even for simple calls like cudaMalloc. The code runs successfully if I restart the computer, but this is far from ideal. This problem is apparently quite common.

My setup is the following:

• OSX 10.6.8
• NVIDIA CUDA drivers : CUDA Driver Version: 4.0.31 (latest)
• GPU Driver Version: 1.6.36.10 (256.00.35f11)

I tried many solutions from the Nvidia forum, but it didn’t work. I don’t want to reboot every time it happens. I also tried to unload and reload the driver with a procedure I assume to be correct (may not be)

kextunload -b com.nvidia.CUDA
kextload -b com.nvidia.CUDA

But still it does not work. How can I kick the GPU (or CUDA) back into sanity ?

This is the device querying result

 CUDA Device Query (Runtime API) version (CUDART static linking)

Found 1 CUDA Capable device(s)

Device 0: "GeForce 9400M"
  CUDA Driver Version / Runtime Version          4.0 / 4.0
  CUDA Capability Major/Minor version number:    1.1
  Total amount of global memory:                 254 MBytes (265945088 bytes)
  ( 2) Multiprocessors x ( 8) CUDA Cores/MP:     16 CUDA Cores
  GPU Clock Speed:                               1.10 GHz
  Memory Clock rate:                             1075.00 Mhz
  Memory Bus Width:                              128-bit
  Max Texture Dimension Size (x,y,z)             1D=(8192), 2D=(65536,32768), 3D=(2048,2048,2048)
  Max Layered Texture Size (dim) x layers        1D=(8192) x 512, 2D=(8192,8192) x 512
  Total amount of constant memory:               65536 bytes
  Total amount of shared memory per block:       16384 bytes
  Total number of registers available per block: 8192
  Warp size:                                     32
  Maximum number of threads per block:           512
  Maximum sizes of each dimension of a block:    512 x 512 x 64
  Maximum sizes of each dimension of a grid:     65535 x 65535 x 1
  Maximum memory pitch:                          2147483647 bytes
  Texture alignment:                             256 bytes
  Concurrent copy and execution:                 No with 0 copy engine(s)
  Run time limit on kernels:                     Yes
  Integrated GPU sharing Host Memory:            Yes
  Support host page-locked memory mapping:       Yes
  Concurrent kernel execution:                   No
  Alignment requirement for Surfaces:            Yes
  Device has ECC support enabled:                No
  Device is using TCC driver mode:               No
  Device supports Unified Addressing (UVA):      No
  Device PCI Bus ID / PCI location ID:           2 / 0
  Compute Mode:
     < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 4.0, CUDA Runtime Version = 4.0, NumDevs = 1, Device = GeForce 9400M
[deviceQuery] test results...
PASSED

This is an example of code that may fail (although in normal conditions it does not)

#include <stdio.h>

__global__ void add(int a, int b, int *c) {
    *c = a + b;
}

int main(void) {
    int c;
    int *dev_c;

    cudaMalloc( (void **) &dev_c, sizeof(int)); // fails here, returning 46

    add<<<1,1>>>(2,7,dev_c);
    cudaMemcpy(&c, dev_c, sizeof(int), cudaMemcpyDeviceToHost);
    printf("hello world, %dn",c);
    cudaFree( dev_c);
    return 0;
}

I also found out that occasionally I get to revert back to a sane behavior without a reboot. I still don’t know what triggers it.

HELP . all CUDA-capable devices are busy or unavailable #65

hey guys i’m having some trouble with the miner i get this errors..

• VERSIONS: XMRig/2.4.2 libuv/1.8.0 CUDA/9.0 gcc/5.4.0
• CPU: Intel(R) Core(TM)2 Quad CPU Q9550 @ 2.83GHz x64 -AES-NI
• GPU #0: GeForce GTX 670 @ 980/3004 MHz 40×21 6×25 arch:30 SMX:7
• GPU

please add the amd version #1 : GeForce GTX 750 Ti @ 1800/2052 MHz 32×8 6×25 arch:21 SMX:4

• VERSIONS: XMRig/2.4.2 libuv/1.8.0 CUDA/9.0 gcc/5.4.0
• CPU: Intel(R) Core(TM)2 Quad CPU Q9550 @ 2.83GHz x64 -AES-NI
• GPU #0: GeForce GTX 670 @ 980/3004 MHz 40×21 6×25 arch:30 SMX:7
• GPU

  please add the amd version #1 : GeForce GTX 750 Ti @ 1800/2052 MHz 32×8 6×25 arch:21 SMX:4

• ALGO: cryptonight, donate=2%
• POOL

  please add the amd version #1 : pool.minexmr.com:7777

• COMMANDS: hashrate, health, pause, resume
  [2017-12-17 18:23:16] use pool pool.minexmr.com:7777 94.23.41.130
  [2017-12-17 18:23:16] new job from pool.minexmr.com:7777 diff 35000
  [CUDA] Error gpu 1: :215 «no kernel image is available for execution on the device»
  terminate called after throwing an instance of ‘std::runtime_error’
  what(): [CUDA] Error: no kernel image is available for execution on the device
  Aborted

they are both cuda related but they are not the same anyone have an idea?

The text was updated successfully, but these errors were encountered:

Источник

RuntimeError: CUDA error: all CUDA-capable devices are busy or unavailable #5598

I’m not trying to use docker I’m trying to use pytorch directly from python
This is the code
«
(base) gegg00@DESKTOP-KS0OLS9:

$ python
Python 3.7.6 (default, Jan 8 2020, 19:59:22)
[GCC 7.3.0] :: Anaconda, Inc. on linux
Type «help», «copyright», «credits» or «license» for more information.

import torch
torch.cuda.is_available()
True
torch.randn(5)
tensor([-2.6408, -1.0831, -1.6984, 0.4742, -0.5909])
torch.randn(5).to(2)
Traceback (most recent call last):
File «», line 1, in
RuntimeError: CUDA error: all CUDA-capable devices are busy or unavailable
«

Environment

Platform ServicePack Version VersionString

Win32NT 10.0.20170.0 Microsoft Windows NT 10.0.20170.0

Ubuntu
Release: 18.04

WSL 2:
Linux version 4.19.121-microsoft-standard (oe-user@oe-host) (gcc version 8.2.0 (GCC)) #1 SMP Fri Jun 19 21:06:10 UTC 2020

Steps to reproduce

I follow the steps
https://docs.nvidia.com/cuda/wsl-user-guide/index.html
Installed Anaconda Anaconda3-2020.02-Linux-x86_64.sh
Installed pytorch
conda install pytorch torchvision cudatoolkit=10.2 -c pytorch
try the code
«
(base) gegg00@DESKTOP-KS0OLS9:

$ python
Python 3.7.6 (default, Jan 8 2020, 19:59:22)
[GCC 7.3.0] :: Anaconda, Inc. on linux
Type «help», «copyright», «credits» or «license» for more information.

import torch
torch.cuda.is_available()
True
torch.randn(5)
tensor([-2.6408, -1.0831, -1.6984, 0.4742, -0.5909])
torch.randn(5).to(2)
Traceback (most recent call last):
File «», line 1, in
RuntimeError: CUDA error: all CUDA-capable devices are busy or unavailable
«
Getting this error:
RuntimeError: CUDA error: all CUDA-capable devices are busy or unavailable

Expected behavior

Actual behavior

The text was updated successfully, but these errors were encountered:

You should try the the new nvidia driver 460.15 and last kernel 4.19.128-microsoft-standard. Both have fixes for CUDA in WSL2.
https://developer.nvidia.com/cuda/wsl/download

I have the same problem and have the same kernel version but nvidia driver 460.20 torch.cuda.is_available() returns true but cannot perform any operation on cuda

I am having the same problem, with NVIDIA driver 460.20, and WSL 2 kernel version 4.19.128.

Same problem. NVIDIA 460.20 and WSL 2 4.19.128. It fails with the NVIDIA Sample for BlackSholes. I have two GPU’s that show up in the Device Query. Same thing with our use of Torch CUDA.

Same problem as well. DeviceQuery runs but not BlackScholes or bandwidthTest , which both error with code=46(cudaErrorDevicesUnavailable) . In pytorch, torch.cuda.is_available() returns True but sending anything to CUDA returns RuntimeError: CUDA error: all CUDA-capable devices are busy or unavailable

I started running into this today with NVIDIA 460.15, WSL 2 4.19.128, Windows 20226.1000. I then updated the driver to NVIDIA 460.20, uninstalled/reinstalled cuda-toolkit-11.1 , restarted windows/wsl, but am still facing the same issues. wsl —status reports that Windows Subsystem for Linux was last updated on 9/7/2020 , while Windows 10 Insider Preview 20226.1000 was updated 10/3/2020. Could the windows update have broke this?

@yhtgrace Here is a quick fix:

Go to Settings -> Recovery -> Click on Get Started under Go back to previous version of Windows 10

This will take 5-10 minutes to roll back to previous version.

Thanks @rohitsingh02! Rolling back to 20221.1000 (while staying at WSL 2 4.19.128, NVIDIA 460.15) resolved the issue for me

Go to Settings -> Recovery -> Click on Get Started under Go back to previous version of Windows 10

This will take 5-10 minutes to roll back to previous version.

It may work. But there was an unfixed issue in version 20221 that caused the following error when running some Linux distros: “The remote procedure call failed”

I guess we just have to wait until Windows gets fixed again.

I got the same problem, but my Windows upgrade from very old version(19042.546).
So when I try to update Windows, it will download Windows 10 Insider Preview 20231.1000, but seem this version not work.
There is any way to upgrade to 20221.1000?

I try to downgrade to 20221 but got another bug.

On WSL 2 / 20231.10005 / Linux 4.19.128-microsoft-standard #1 SMP Tue Jun 23 12:58:10 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux

Any other workaround than 20221? I only just hopped on WI for WSL2 / CUDA support in the last day so I think I went straight to 20231 from 1903 so not sure I can even easily downgrade.

If it helps, what I chose to do as a workaround is simply run my CUDA code on the Windows command line as opposed to WSL. Of course, this requires that you reproduce whatever setup you had in WSL, e.g. installing packages. It’s not really a solution, because you’re just using a different OS, but hopefully it helps you do your work 🙂

This should be fixed on build 20236.

If it helps, what I chose to do as a workaround is simply run my CUDA code on the Windows command line as opposed to WSL. Of course, this requires that you reproduce whatever setup you had in WSL, e.g. installing packages. It’s not really a solution, because you’re just using a different OS, but hopefully it helps you do your work 🙂

Unfortunately I was trying to use the fastai lib which specifically really doesn’t support Windows.

Build 20236 seems to work:

Thanks guys! Now to just get fastai working..heh

Источник

Cuda devices unavailable or busy #73

out of 15 jobs usually 2-5 give me the following error message:

THCudaCheck FAIL file=/pytorch/aten/src/THC/THCTensorRandom.cu line=25 error=46 : all CUDA-capable devices are busy or unavailable

It would be very nice if you could fix this.
Since it happens so often I do not think this is related to any specific node, but a general problem.

However, if you need more information in order to tackle the issue, please let me know!

The text was updated successfully, but these errors were encountered:

This seems to be a common issue caused by the tensorflow docker container not releasing GPUs when slurm kills the job thus leaving slurm believing a GPU is free but it is in fact still locked by a process that hasn’t been correctly reaped.

This can also happen when someone manually overrides CUDA_VISIBLE_DEVICES themselves to take over free, but unallocated (by slurm) GPUs on a node. I don’t think there are any permission solutions in place to prevent this sort of abuse.

Is it possible to check the gpu memory usage after a job is finished or killed?

Same proble slurmstepd-dgj401: error: task_p_post_term: rmdir(/dev/cpuset/slurm_dgj401_304107/slurm304107.4294967294_0) failed Device or resource busy

I confirm to also get these errors

I obtain the same error, and it is occurring quite frequently these days.

For instance on a single-GPU job where I am assigned GPU 1:

One can see that device 1 has no process running on it, yet all the memory is already used.

It is possible to find the processes responsible for locking the GPU memory with a command like:
lsof /dev/nvidia$CUDA_VISIBLE_DEVICES (https://stackoverflow.com/a/4571250).

Using this command, maybe the processes listed by lsof could be killed at either the start or end of each job, to make sure that the GPU memory is effectively free?

Источник

all CUDA-capable devices are busy or unavailable (only with 2 gpus) #319

2 GPUs (gtx760 and gtx460)
Arch Linux

I compiled xmr-stak with CUDA 8 and gcc-5 since I needed the ability to use compute20/21 for my gtx 460. I just exported CC and CXX, make . make install.

I configured my nvidia.txt this way

«gpu_threads_conf» :
[
// gpu: GeForce GTX 760 architecture: 30
// memory: 1920/1991 MiB
// smx: 6
< «index» : 0,
«threads» : 80, «blocks» : 12,
«bfactor» : 0, «bsleep» : 0,
«affine_to_cpu» : false,
>,
// gpu: GeForce GTX 460 architecture: 21
// memory: 816/964 MiB
// smx: 6
< «index» : 1,
«threads» : 34, «blocks» : 12,
«bfactor» : 0, «bsleep» : 0,
«affine_to_cpu» : false,
>,
],

when I run it I get this problem

[2017-12-02 01:43:47] : MEMORY ALLOC FAILED: mmap failed
[2017-12-02 01:43:47] : MEMORY ALLOC FAILED: mmap failed

Brought to you by fireice_uk and psychocrypt under GPLv3.
Based on CPU mining code by wolf9466 (heavily optimized by fireice_uk).
Based on NVIDIA mining code by KlausT and psychocrypt.
Based on OpenCL mining code by wolf9466.

Configurable dev donation level is set to 2.0%

You can use following keys to display reports:
‘h’ — hashrate
‘r’ — results
‘c’ — connection

[2017-12-02 01:43:47] : Start mining: MONERO
[2017-12-02 01:43:47] : Starting NVIDIA GPU thread 0, no affinity.
[2017-12-02 01:43:47] : Starting NVIDIA GPU thread 1, no affinity.
[2017-12-02 01:43:47] : MEMORY ALLOC FAILED: mmap failed
[2017-12-02 01:43:47] : MEMORY ALLOC FAILED: mmap failed
[2017-12-02 01:43:47] : WARNING: No AMD OpenCL platform found. Possible driver issues or wrong vendor driver.
[2017-12-02 01:43:47] : WARNING: backend AMD disabled.
[2017-12-02 01:43:47] : Starting single thread, affinity: 0.
[CUDA] Error gpu 0: :206
terminate called after throwing an instance of ‘std::runtime_error’
what(): [CUDA] Error: all CUDA-capable devices are busy or unavailable
[1] 12493 abort (core dumped) ./xmr-stak/build/bin/xmr-stak

I know this settings work because I can comment out one card or the other and they work just fine alone but they fail when I run them together.

This is the output of nvidia-smi

Any ideas of what could be causing this?
Thanks

The text was updated successfully, but these errors were encountered:

Why would the number of threads matter if each card is able to handle that load alone? Do I have to calculate threads and blocks differently if both cards work at the same time?

I might be missing something. The error doesn’t seem to be an out of memory one. Each card is able to mine with those threads just fine as long as I don’t run them together. Unless you are talking about some other memory instead of vram.

Do you mind elaborating why 2 cards that work individually at those threads give an «what(): [CUDA] Error: all CUDA-capable devices are busy or unavailable» when I try to launch them together? Even without xorg running.

I just compiled the dev branch and I still get the same error. Out of curiosity I reduced the threads to under 10 on both cards and that didn’t have any effect on the outcome either. Both cards are still able to work with those settings as long as I have one of them commented out.

I also tried to run two separate instances of xmr-stak with different configs (one for each card) but I always get

WARNING: NVIDIA Unable to query number of CUDA devices!
WARNING: NVIDIA no device found
[2017-12-02 03:44:25] : WARNING: backend NVIDIA disabled.
[2017-12-02 03:44:25] : WARNING: UNKNOWN_ERROR when calling clGetPlatformIDs for number of platforms.

when I try to launch the second instance. I assume this is normal but I wanted to try anyway.

[2017-12-02 03:50:31] : MEMORY ALLOC FAILED: mmap failed
[2017-12-02 03:50:31] : MEMORY ALLOC FAILED: mmap failed

Brought to you by fireice_uk and psychocrypt under GPLv3.
Based on CPU mining code by wolf9466 (heavily optimized by fireice_uk).
Based on NVIDIA mining code by KlausT and psychocrypt.
Based on OpenCL mining code by wolf9466.

Configurable dev donation level is set to 2.0%

You can use following keys to display reports:
‘h’ — hashrate
‘r’ — results
‘c’ — connection

[2017-12-02 03:50:31] : Start mining: MONERO
[2017-12-02 03:50:31] : NVIDIA: GPU configuration stored in file ‘nvidia.txt’
[2017-12-02 03:50:31] : Starting NVIDIA GPU thread 0, no affinity.
[2017-12-02 03:50:31] : Starting NVIDIA GPU thread 1, no affinity.
[2017-12-02 03:50:31] : MEMORY ALLOC FAILED: mmap failed
[2017-12-02 03:50:31] : MEMORY ALLOC FAILED: mmap failed
[2017-12-02 03:50:31] : WARNING: No AMD OpenCL platform found. Possible driver issues or wrong vendor driver.
[2017-12-02 03:50:31] : WARNING: backend AMD disabled.
[2017-12-02 03:50:31] : Starting single thread, affinity: 0.
[CUDA] Error gpu 0: :206
terminate called after throwing an instance of ‘std::runtime_error’
what(): [CUDA] Error: all CUDA-capable devices are busy or unavailable
[1] 6695 abort (core dumped) ./xmr-stak-dev/build/bin/xmr-stak

This is the nvidia.txt generated (without the comments)

«gpu_threads_conf» :
[
// gpu: GeForce GTX 760 architecture: 30
// memory: 1529/1991 MiB
// smx: 6
< «index» : 0,
«threads» : 38, «blocks» : 18,
«bfactor» : 0, «bsleep» : 0,
«affine_to_cpu» : false,
>,
// gpu: GeForce GTX 460 architecture: 21
// memory: 886/964 MiB
// smx: 7
< «index» : 1,
«threads» : 26, «blocks» : 14,
«bfactor» : 0, «bsleep» : 0,
«affine_to_cpu» : false,
>,

Is there anything else I can do to provide more info on this issue? Any clues if it’s a problem on my end or the miner?

Sorry to bother, not being able to mine with one of the cards feels awful.

I have been doing this already. Somewhere else on the documentation I was reading about large pages and even though I didn’t read carefully (at the time I was figuring how to compile with cuda8 and gcc5) I did read that launching the script with sudo would avoid some of those issues. At least until I configure that properly.

Is there a log file I could access that would be a little more verbose about the error?

I doubt there will be anything here but this are the last entries of my dmesg. Should I normally be able to launch 2 separate instances of the script with different configs that is targeted to each card? When I try to do that I get a different error but I’m not sure if that is even normally possible.

Источник

On WSL 2 / 20231.10005 / Linux 4.19.128-microsoft-standard #1 SMP Tue Jun 23 12:58:10 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux

(fastai) forbiddenera@forbiddenera:/usr/local/cuda/samples/bin/x86_64/linux/release$ ./deviceQueryDrv
./deviceQueryDrv Starting...

CUDA Device Query (Driver API) statically linked version
Detected 1 CUDA Capable device(s)

Device 0: "GeForce RTX 2070 SUPER"
  CUDA Driver Version:                           11.2
  CUDA Capability Major/Minor version number:    7.5
  Total amount of global memory:                 8192 MBytes (8589934592 bytes)
  (40) Multiprocessors, ( 64) CUDA Cores/MP:     2560 CUDA Cores
  GPU Max Clock rate:                            1905 MHz (1.90 GHz)
  Memory Clock rate:                             7001 Mhz
  Memory Bus Width:                              256-bit
  L2 Cache Size:                                 4194304 bytes
  Max Texture Dimension Sizes                    1D=(131072) 2D=(131072, 65536) 3D=(16384, 16384, 16384)
  Maximum Layered 1D Texture Size, (num) layers  1D=(32768), 2048 layers
  Maximum Layered 2D Texture Size, (num) layers  2D=(32768, 32768), 2048 layers
  Total amount of constant memory:               65536 bytes
  Total amount of shared memory per block:       49152 bytes
  Total number of registers available per block: 65536
  Warp size:                                     32
  Maximum number of threads per multiprocessor:  1024
  Maximum number of threads per block:           1024
  Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
  Max dimension size of a grid size (x,y,z):    (2147483647, 65535, 65535)
  Texture alignment:                             512 bytes
  Maximum memory pitch:                          2147483647 bytes
  Concurrent copy and kernel execution:          Yes with 2 copy engine(s)
  Run time limit on kernels:                     Yes
  Integrated GPU sharing Host Memory:            No
  Support host page-locked memory mapping:       Yes
  Concurrent kernel execution:                   Yes
  Alignment requirement for Surfaces:            Yes
  Device has ECC support:                        Disabled
  Device supports Unified Addressing (UVA):      Yes
  Device supports Managed Memory:                Yes
  Device supports Compute Preemption:            Yes
  Supports Cooperative Kernel Launch:            Yes
  Supports MultiDevice Co-op Kernel Launch:      Yes
  Device PCI Domain ID / Bus ID / location ID:   0 / 1 / 0
  Compute Mode:
     < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
Result = PASS
(fastai) forbiddenera@forbiddenera:/usr/local/cuda/samples/bin/x86_64/linux/release$ ./convolutionFFT2D
[./convolutionFFT2D] - Starting...
GPU Device 0: "Turing" with compute capability 7.5

Testing built-in R2C / C2R FFT-based convolution
...allocating memory
CUDA error at main.cpp:118 code=46(cudaErrorDevicesUnavailable) "cudaMalloc((void **)&d_Data, dataH * dataW * sizeof(float))"

Any other workaround than 20221? I only just hopped on WI for WSL2 / CUDA support in the last day so I think I went straight to 20231 from 1903 so not sure I can even easily downgrade.

I am trying to run CUDA 7.5 toolkit in an Ubuntu 16.04 LTS environment and have gotten very close to having the full capabilities. I can run the deviceQuerry as shown below, but I can not run certain samples as is also shown below (nbody).

I was having an issue with the libGLU.so library and associated files but I was able to get past those errors… it is now that I am having no luck solving the error listed below…

Do you know what might be the problem and what I could try to fix this?

___@___:/usr/local/cuda/samples/1_Utilities/deviceQuery$ ./deviceQuery
./deviceQuery Starting...

 CUDA Device Query (Runtime API) version (CUDART static linking)

Detected 1 CUDA Capable device(s)

Device 0: "Quadro K620"
  CUDA Driver Version / Runtime Version          8.0 / 7.5
  CUDA Capability Major/Minor version number:    5.0
  Total amount of global memory:                 1993 MBytes (2090008576 bytes)
  ( 3) Multiprocessors, (128) CUDA Cores/MP:     384 CUDA Cores
  GPU Max Clock rate:                            1124 MHz (1.12 GHz)
  Memory Clock rate:                             900 Mhz
  Memory Bus Width:                              128-bit
  L2 Cache Size:                                 2097152 bytes
  Maximum Texture Dimension Size (x,y,z)         1D=(65536), 2D=(65536, 65536), 3D=(4096, 4096, 4096)
  Maximum Layered 1D Texture Size, (num) layers  1D=(16384), 2048 layers
  Maximum Layered 2D Texture Size, (num) layers  2D=(16384, 16384), 2048 layers
  Total amount of constant memory:               65536 bytes
  Total amount of shared memory per block:       49152 bytes
  Total number of registers available per block: 65536
  Warp size:                                     32
  Maximum number of threads per multiprocessor:  2048
  Maximum number of threads per block:           1024
  Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
  Max dimension size of a grid size    (x,y,z): (2147483647, 65535, 65535)
  Maximum memory pitch:                          2147483647 bytes
  Texture alignment:                             512 bytes
  Concurrent copy and kernel execution:          Yes with 1 copy engine(s)
  Run time limit on kernels:                     Yes
  Integrated GPU sharing Host Memory:            No
  Support host page-locked memory mapping:       Yes
  Alignment requirement for Surfaces:            Yes
  Device has ECC support:                        Disabled
  Device supports Unified Addressing (UVA):      Yes
  Device PCI Domain ID / Bus ID / location ID:   0 / 1 / 0
  Compute Mode:
     < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 8.0, CUDA Runtime Version = 7.5, NumDevs = 1, Device0 = Quadro K620
Result = PASS



___@___:/usr/local/cuda/samples/5_Simulations/nbody$ ./nbody
Run "nbody -benchmark [-numbodies=<numBodies>]" to measure performance.
    -fullscreen       (run n-body simulation in fullscreen mode)
    -fp64             (use double precision floating point values for simulation)
    -hostmem          (stores simulation data in host memory)
    -benchmark        (run benchmark to measure performance) 
    -numbodies=<N>    (number of bodies (>= 1) to run in simulation) 
    -device=<d>       (where d=0,1,2.... for the CUDA device to use)
    -numdevices=<i>   (where i=(number of CUDA devices > 0) to use for simulation)
    -compare          (compares simulation results running once on the default GPU and once on the CPU)
    -cpu              (run n-body simulation on the CPU)
    -tipsy=<file.bin> (load a tipsy model file for simulation)

NOTE: The CUDA Samples are not meant for performance measurements. Results may vary when GPU Boost is enabled.

> Windowed mode
> Simulation data stored in video memory
> Single precision floating point simulation
> 1 Devices used for simulation
> Compute 5.0 CUDA device: [Quadro K620]
CUDA error at bodysystemcuda_impl.h:160 code=46(cudaErrorDevicesUnavailable) "cudaEventCreate(&m_deviceData[0].event)"