04:21:34.676043: I tensorflow/stream_executor/platform/default/dso_:48] Successfully opened dynamic library libcusolver.so.10 04:21:34.673378: I tensorflow/stream_executor/platform/default/dso_:48] Successfully opened dynamic library libcurand.so.10 04:21:34.672971: I tensorflow/stream_executor/platform/default/dso_:48] Successfully opened dynamic library libcufft.so.10 04:21:34.670275: I tensorflow/stream_executor/platform/default/dso_:48] Successfully opened dynamic library libcublas.so.10 04:21:34.667644: I tensorflow/stream_executor/platform/default/dso_:48] Successfully opened dynamic library libcudart.so.10.1 04:21:34.667613: I tensorflow/core/common_runtime/gpu/gpu_:1716] Found device 2 with properties: 04:21:34.667064: I tensorflow/stream_executor/cuda/cuda_gpu_:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero 04:21:34.667017: I tensorflow/core/common_runtime/gpu/gpu_:1716] Found device 1 with properties: 04:21:34.666459: I tensorflow/stream_executor/cuda/cuda_gpu_:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero PciBusID: 0000:00:0c.0 name: Tesla V100-SXM2-16GB computeCapability: 7.0ĬoreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.75GiB deviceMemoryBandwidth: 836.37GiB/s
04:21:34.666369: I tensorflow/core/common_runtime/gpu/gpu_:1716] Found device 0 with properties: 04:21:34.665738: I tensorflow/stream_executor/cuda/cuda_gpu_:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero 04:21:34.499432: I tensorflow/stream_executor/platform/default/dso_:48] Successfully opened dynamic library libcuda.so.1 In : print("Num GPUs Available: ", len(tf.config.list_physical_devices('GPU'))) 04:21:31.110443: I tensorflow/stream_executor/platform/default/dso_:48] Successfully opened dynamic library libcudart.so.10.1 The following test is good: import tensorflow as tf My conda env list shows: cudatoolkit 10.1.243 h6bb024c_0
#CUDA UPDATE MAC DRIVER#
I installed cudatoolkit=10.1, but the cuda driver still not good.
#CUDA UPDATE MAC HOW TO#
If I can't keep cuda 10.0, how to directly upgrade cuda to 10.1 with or without conda? It's best if I can upgrade in Conda.
#CUDA UPDATE MAC INSTALL#
If I want to keep the old version cuda 10.0, can I update cuda to 10.1 through Conda? This won't work: conda install cuda=10.1
I know how to install cudakit in conda: conda install cudatoolkit=10.1īut this seems not enough: Status: CUDA driver version is insufficient for CUDA runtime version I want to install Tensorflow 2.3/2.4, so I need to upgrade cuda to 10.1 at least in Conda. | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr.