CUDA runtime implicit initialization on GPU:0 failed. Status: out of memory

[jayaramreddy10]

Hi,

When i test npcr_ScanNet.py using pretrained checkpoints , I am getting below error message that GPU is out of memory. But I am already using a high end GPU- NVIDIA RTX 2070, is it not sufficient for inference?

Error message: load ply time: 25.08699655532837 loaded descriptors. Traceback (most recent call last): File "npcr_ScanNet.py", line 69, in sess = tf.Session() File "/scratch/jayaram.messi/envs/Neural_pc_rendering_tf/lib/python3.6/site-packages/tensorflow/python/client/session.py", line 1551, in init
super(Session, self).init(target, graph, config=config) File "/scratch/jayaram.messi/envs/Neural_pc_rendering_tf/lib/python3.6/site-packages/tensorflow/python/client/session.py", line 676, in init
self._session = tf_session.TF_NewSessionRef(self._graph._c_graph, opts) tensorflow.python.framework.errors_impl.InternalError: CUDA runtime implicit initialization on GPU:0 failed. Status: out of memory

So, how do I avoid this out of memory issue ?

[daipengwa]

Hi, We used 1080Ti, there is no OOM error. I suggest that you can check if the GPU is occupied by other processes? Thanks

On Mon, Feb 27, 2023 at 10:45 PM jayaramreddy10 @.***> wrote:

Hi,

When i test npcr_ScanNet.py using pretrained checkpoints , I am getting below error message that GPU is out of memory. But I am already using a high end GPU- NVIDIA RTX 2070, is it not sufficient for inference?

Error message: load ply time: 25.08699655532837 loaded descriptors. Traceback (most recent call last): File "npcr_ScanNet.py", line 69, in sess = tf.Session() File "/scratch/jayaram.messi/envs/Neural_pc_rendering_tf/lib/python3.6/site-packages/tensorflow/python/client/session.py", line 1551, in init super(Session, self).init(target, graph, config=config) File "/scratch/jayaram.messi/envs/Neural_pc_rendering_tf/lib/python3.6/site-packages/tensorflow/python/client/session.py", line 676, in init self._session = tf_session.TF_NewSessionRef(self._graph._c_graph, opts) tensorflow.python.framework.errors_impl.InternalError: CUDA runtime implicit initialization on GPU:0 failed. Status: out of memory

So, how do I avoid this out of memory issue ?

— Reply to this email directly, view it on GitHub https://github.com/daipengwa/Neural-Point-Cloud-Rendering-via-Multi-Plane-Projection/issues/5, or unsubscribe https://github.com/notifications/unsubscribe-auth/AH2JWYKG2TD57MAWARWSBILWZS4YBANCNFSM6AAAAAAVJP2VEA . You are receiving this because you are subscribed to this thread.Message ID: <daipengwa/Neural-Point-Cloud-Rendering-via-Multi-Plane-Projection/issues/5 @github.com>

[jayaramreddy10]

Thanks, its resolved now