Solvedcaffe nvcc fatal : Unsupported gpu architecture 'compute_61'

Issue summary

I have the following configuration below but at the very start of compilation, it throws the errors for all gpu modules:

nvcc fatal : Unsupported gpu architecture 'compute_61'

Why is this the case? Is 7.5 not supported on the master branch for the new Geforce 1070/1080 cards?

Your system configuration

Operating system:
Compiler: GCC 4.8.4
CUDA version (if applicable): 7.5
CUDNN version (if applicable): 5.0
Python or MATLAB version (for pycaffe and matcaffe respectively):

28 Answers

✔️Accepted Answer

You should modify Makefile.config.

\# CUDA architecture setting: going with all of them.
\# For CUDA < 6.0, comment the *_50 through *_61 lines for compatibility.
\# For CUDA < 8.0, comment the *_60 and *_61 lines for compatibility.
CUDA_ARCH := -gencode arch=compute_20,code=sm_20 \
                -gencode arch=compute_20,code=sm_21 \
                -gencode arch=compute_30,code=sm_30 \
                -gencode arch=compute_35,code=sm_35 \
                -gencode arch=compute_50,code=sm_50 \
                -gencode arch=compute_52,code=sm_52
\#               -gencode arch=compute_60,code=sm_60 \
\#               -gencode arch=compute_61,code=sm_61 \
\#               -gencode arch=compute_61,code=compute_61

Other Answers:

Hello Everyone,

I am using CUDA 9.0 on Ubuntu 16.04 and I am facing the following error:

nvcc fatal   : Unsupported gpu architecture 'compute_20'

This can be resolved by removing the lines in the Makefile.config :

-gencode arch=compute_20,code=sm_20 \
 -gencode arch=compute_20,code=sm_21 \


Where can i find the Makefile.config? I have the same problem, but i can't find this archive to edit.

Great! Thank you very much, @naibaf7, I am a bit new to compiling stuff and I still get confused with the differences between Make and CMake. I'll research it for future issues.

So following your response, I modified the cmake/Cuda.cmake file and removed the architectures 60 and 61 in lines 7 and 92:93. Now everything seems to be working perfectly.

Thanks again!!

Related Issues:

caffe version `GLIBCXX_3.4.21' not found
^ Should fix this error Please ask installation questions only in the Caffe mailing list. ...
caffe enabled -std=c++11
@manifoldQAQ Thanks this worked for me I also had to add it to the NVCCFLAGS in the makefile for it ...
caffe nvcc fatal : Unsupported gpu architecture 'compute_61'
You should modify Makefile.config. Issue summary I have the following configuration below but at the...
caffe .build_release/lib/ undefined reference to `cv::imdecode
Hi @monajalal Have you noticed the 21st line in Makefile.config.example which is # OPENCV_VERSION :=...
caffe Error in make mattest
@cjjsaic @yl1991 ubuntu 16.04 matlab 2017a cuda 8.0 g++/gcc 5.4 Issue summary I'm trying to compile ...
caffe caffe build problem
I think the reason is you use make to compile which makes caffe's python port only find libraries in...
caffe problem.. /usr/bin/ld: cannot find -lopencv_imgcodecs
opencv_imgcodecs is needed and available in opencv 3.0+ You don't need to link it if using version 2...
caffe malloc error on macOS Sierra 10.12
the problem is related to leveldb If leveldb is installed with brew by default it depends on the goo...
caffe problem
On CentOS is this a problem? ImportError: dlopen(caffe/ 2): Library not loaded:
caffe when can caffe support visual studio 2017
@mjahanifar If you want to help can you try my vcpkg port: With VS 2017 and CUDA 9 installed open a ...
keras Loading model with custom loss function: ValueError: 'Unknown loss function'
I solved this problem by adding 'custom_bojects' my loss function: I trained and saved a model that ...
keras Tensorflow backend - bug in model._make_predict_function(...)
I had this problem when doing inference in a different thread than where I loaded my model ...
keras Creating Model variable throws "AttributeError: 'Tensor' object has no attribute '_keras_history'"
@zaid478 oh I assumed you were doing it on coursera Anyway Hello all setup: Keras 2.0 Tensorflow 1.0...
tensorflow ImportError: cannot open shared object file: No such file or directory
@rhinsall I just found this out myself not sure if it's common knowledge but got around this by doin...
keras "<tensor> is not an element of this graph." when loading model.
I had a problem similar to that of @piraka9011 which was solved by calling model._make_predict_funct...
pytorch I can't import PyTorch, libomp.dylib can't be loaded.
brew install libomp solves the problem. 🐛 Bug I tried to install PyTorch but I can't use it ...
pytorch RuntimeError: cuda runtime error (59) : device-side assert triggered when running transfer_learning_tutorial
@soumith Hi I find the error it is due to my labels contain value -1 which is unacceptable in criter...
tensorflow At Runtime : "Error while reading resource variable softmax/kernel from Container: localhost"
I had the same issue in tensorflow 1.13.1 which I have resolved by creating a reference to the sessi...
tensorflow Mac + Python 3.6.1: Attempting to download mnist data results in CERTIFICATE_VERIFY_FAILED error
This might have also been because I recently switched to a new machine and it's still only partially...
tensorflow UnrecognizedFlagError: Unknown command line flag 'f'
I solved this issue by adding the line:'f' '' 'kernel') what this essent...
tensorflow FutureWarning: Deprecated numpy API calls in tf.python.framework.dtypes
@Akeaakar And for the time being pip install numpy<1.17 to revert to numpy version 1.16.4 ...
tensorflow ERROR: Cannot uninstall 'wrapt'. during upgrade
This worked for me: pip install wrapt --upgrade --ignore-installed pip install tensorflow ...
pytorch ModuleNotFoundError: No module named 'torch._C'
Solution the problem is that you have a folder called torch in the same directory which is being pic...
tensorflow Keras application - Tensor is not an element of this graph on eval after train
K.clear_session() did not work for me however what worked was : While predicting use the same graph ...
keras Quick Question: can a model be fit for multiple times?
Yes successive calls to fit will incrementally train the model. I'm now having something around 5000...
tensorflow InternalError: Blas GEMM launch failed
Make sure you have no other processes using the GPU running Run nvidia-smi to check this. ...
pytorch RuntimeError: DataLoader worker is killed by signal: Killed.
I've encountered the same problem recently If you're using the docker to run the PyTorch program wit...
tensorflow Is python 3.7.x supported with Tensorflow
I sill get Ive been trying to install Tensorflow on my computer which currently runs python 3.7 howe...
tensorflow Tensorflow installation Error in Mac Sierra due to numpy
Also Had to do this with the six dependency as well Hi I want to install TensorFlow on Mac Currently...
keras Error when checking model target: expected activation_2 to have shape (None, 10) but got array with shape (3, 1)
Your model has an output of shape (10,) however your outputs have dimension (1,) You probably want t...
tensorflow tensorboard: view graph from saved_model.pb file [feature request]
@brandondutra As far as I know you don't need to create any summaries to load the graph into Tensorb...
keras Is there a way in Keras to apply different weights to a cost function?
Ok so I had the time to quickly test it Hi there I am trying to implement a classification problem w...
tensorflow ImportError: cannot open shared object file: No such file or directory
Double check that you have exported the LD_LIBRARY_PATH export LD_LIBRARY_PATH=/usr/local/cuda-8.0/l...
pytorch [PyTorch][Feature Request] Label Smoothing for CrossEntropyLoss
Here is my implement Hi guys The type torch.LongTensor of target will hinder the implementation like...
keras Keras model cannot be loaded if it contains a Lambda layer calling tf.image.resize_images
Try using custom_objects argument in load_model (or model_from_json) function For your case load_mod...
keras how to use fit_generator with multiple image inputs
Thanks Here is how I solved it following some ideas from issue 3386 maybe someone might find it usef...
tensorflow Tensorflow Failed to create Session
maybe out of GPU memory? Try running with export CUDA_VISIBLE_DEVICES='' Hi I tried basic program in...
keras Expected 3 dimensions but got array with shape (11, 2)
The problem is that you start with a three dimensional layer but never reduce the dimensionality in ...
keras AttributeError: 'ProgbarLogger' object has no attribute 'log_values'
This happens if steps_per_epoch is 0 Make sure that your batch size is not greater than the dataset ...
keras For large datasets, which to use: fit or train_on_batch?
I think the best way to fit large data which can not fit into memory is writing a customize generato...
keras Split train data into training and validation when using ImageDataGenerator and model.fit_generator
I just found this in the documentation given here Its okay if I am keeping my training and validatio...
pytorch optimizer load_state_dict() problem?
@apaszke Ah my bad I forgot to update the line where the optimizer is recreated But otherwise the fo...
keras using pre trained VGG16 for another classification task
One way to do this is to not include the fully-connected layers at the top of the network Then add n...
tensorflow Error in `python': double free or corruption (!prev)
Can confirm this I ran into this on my CI server and the the following fixed it: ...
tensorflow Why this error on tensorflow 1.13.1 with python 2.7 : ImportError: No module named model_utils
I was able to solve this under Python 3.6 on Linux with: System information Have I written custom co...
keras "Merge" versus "merge", what is the difference?
Merge is a layer Merge takes layers as input Merge is usually used with Sequential models merge is a...
tensorflow TF 1.x: remove the "deprecated" warning messages
On 1.14.0 only this works: I know the functional APIs such as tf.layers.dense will disappear in TF 2...