Different order between nvidia-smi and nvidia x server settings
up vote
-1
down vote
favorite
When I run the command nvidia-smi
, I get the following two GPUs ordered by their Bus-Ids:
For GPU 0, 00000000:0A:00.0
For GPU 1, 00000000:41:00.0
However, when I run NVIDIA X server Settings
, I can get the following GPUs ordered by their Bus-Ids:
For GPU 0, PCI:65:0:0
For GPU 1, PCI:10:0:0
Thus, they have different orders based on their own ways to enumerate Bus-Ids.
Is there any way to make the orders consistent?
OS: Ubuntu 16.04
Mainboard: MSI X399 (for AMD 1950X)
(PS)
In fact, my computer is freezed when using the dataparallel mechanism for deep learning on multiple GPUs. I guess the different GPU orders are the reason.
ubuntu cuda gpu nvidia pci-e
add a comment |
up vote
-1
down vote
favorite
When I run the command nvidia-smi
, I get the following two GPUs ordered by their Bus-Ids:
For GPU 0, 00000000:0A:00.0
For GPU 1, 00000000:41:00.0
However, when I run NVIDIA X server Settings
, I can get the following GPUs ordered by their Bus-Ids:
For GPU 0, PCI:65:0:0
For GPU 1, PCI:10:0:0
Thus, they have different orders based on their own ways to enumerate Bus-Ids.
Is there any way to make the orders consistent?
OS: Ubuntu 16.04
Mainboard: MSI X399 (for AMD 1950X)
(PS)
In fact, my computer is freezed when using the dataparallel mechanism for deep learning on multiple GPUs. I guess the different GPU orders are the reason.
ubuntu cuda gpu nvidia pci-e
add a comment |
up vote
-1
down vote
favorite
up vote
-1
down vote
favorite
When I run the command nvidia-smi
, I get the following two GPUs ordered by their Bus-Ids:
For GPU 0, 00000000:0A:00.0
For GPU 1, 00000000:41:00.0
However, when I run NVIDIA X server Settings
, I can get the following GPUs ordered by their Bus-Ids:
For GPU 0, PCI:65:0:0
For GPU 1, PCI:10:0:0
Thus, they have different orders based on their own ways to enumerate Bus-Ids.
Is there any way to make the orders consistent?
OS: Ubuntu 16.04
Mainboard: MSI X399 (for AMD 1950X)
(PS)
In fact, my computer is freezed when using the dataparallel mechanism for deep learning on multiple GPUs. I guess the different GPU orders are the reason.
ubuntu cuda gpu nvidia pci-e
When I run the command nvidia-smi
, I get the following two GPUs ordered by their Bus-Ids:
For GPU 0, 00000000:0A:00.0
For GPU 1, 00000000:41:00.0
However, when I run NVIDIA X server Settings
, I can get the following GPUs ordered by their Bus-Ids:
For GPU 0, PCI:65:0:0
For GPU 1, PCI:10:0:0
Thus, they have different orders based on their own ways to enumerate Bus-Ids.
Is there any way to make the orders consistent?
OS: Ubuntu 16.04
Mainboard: MSI X399 (for AMD 1950X)
(PS)
In fact, my computer is freezed when using the dataparallel mechanism for deep learning on multiple GPUs. I guess the different GPU orders are the reason.
ubuntu cuda gpu nvidia pci-e
ubuntu cuda gpu nvidia pci-e
asked Nov 9 at 13:17
hjung
224
224
add a comment |
add a comment |
1 Answer
1
active
oldest
votes
up vote
1
down vote
The order that matters is CUDA order.
You can enumerate CUDA devices and see which one is which.
Then, you'll be able to run your program on the GPU that is not being used for display using CUDA_VISIBLE_DEVICES
, cudaSetDevice()
or cuCtxCreate()
add a comment |
1 Answer
1
active
oldest
votes
1 Answer
1
active
oldest
votes
active
oldest
votes
active
oldest
votes
up vote
1
down vote
The order that matters is CUDA order.
You can enumerate CUDA devices and see which one is which.
Then, you'll be able to run your program on the GPU that is not being used for display using CUDA_VISIBLE_DEVICES
, cudaSetDevice()
or cuCtxCreate()
add a comment |
up vote
1
down vote
The order that matters is CUDA order.
You can enumerate CUDA devices and see which one is which.
Then, you'll be able to run your program on the GPU that is not being used for display using CUDA_VISIBLE_DEVICES
, cudaSetDevice()
or cuCtxCreate()
add a comment |
up vote
1
down vote
up vote
1
down vote
The order that matters is CUDA order.
You can enumerate CUDA devices and see which one is which.
Then, you'll be able to run your program on the GPU that is not being used for display using CUDA_VISIBLE_DEVICES
, cudaSetDevice()
or cuCtxCreate()
The order that matters is CUDA order.
You can enumerate CUDA devices and see which one is which.
Then, you'll be able to run your program on the GPU that is not being used for display using CUDA_VISIBLE_DEVICES
, cudaSetDevice()
or cuCtxCreate()
answered Nov 9 at 13:53
Robin Thoni
744518
744518
add a comment |
add a comment |
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fstackoverflow.com%2fquestions%2f53226454%2fdifferent-order-between-nvidia-smi-and-nvidia-x-server-settings%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown