未检测到 GPU 名称 张量流 2.2.0

Question

我将张量流更新到2.2.0并相应地

nvcc - 11.0
cudnn - 11.0
GPU - GTX 1050 ti

运行以下代码时

print(device_lib.list_local_devices())

我面临以下输出

[name: "/device:CPU:0"
device_type: "CPU"
memory_limit: 268435456
locality {
}
incarnation: 12436950237915670665
, name: "/device:XLA_CPU:0"
device_type: "XLA_CPU"
memory_limit: 17179869184
locality {
}
incarnation: 11900640710651469327
physical_device_desc: "device: XLA_CPU device"
, name: "/device:XLA_GPU:0"
device_type: "XLA_GPU"
memory_limit: 17179869184
locality {
}
incarnation: 6061376473165052950
physical_device_desc: "device: XLA_GPU device"
]

我在任何地方都看不到gtx 1050ti列出，尽管我可以看到设备 0 提到的 GPU，可能意味着英特尔的内置 GPU。

1050ti在张量流和 Cuda 方面的兼容版本是什么？

更新

我尝试了以下命令

print(tf.config.list_physical_devices('GPU'))

结果为空。这是否意味着未检测到 GPU？

Answer 1

此链接提供有关 Cuda、tensorflow 和 Cudnn 兼容版本的信息，以检测 gpu。对于版本张量流 2.2.0、Cudnn7.4和Cuda 10.1、python 3.8，我能够获得以下输出

[name: "/device:CPU:0"
device_type: "CPU"
memory_limit: 268435456
locality {
}
incarnation: 4549764507052008926
, name: "/device:XLA_CPU:0"
device_type: "XLA_CPU"
memory_limit: 17179869184
locality {
}
incarnation: 5130440468361087955
physical_device_desc: "device: XLA_CPU device"
, name: "/device:GPU:0"
device_type: "GPU"
memory_limit: 3136264601
locality {
bus_id: 1
links {
}
}
incarnation: 8742529146709444949
physical_device_desc: "device: 0, name: GeForce GTX 1050 Ti, pci bus id: 
0000:01:00.0, compute capability: 6.1"
, name: "/device:XLA_GPU:0"
device_type: "XLA_GPU"
memory_limit: 17179869184
locality {
}
incarnation: 12774508348529661585
physical_device_desc: "device: XLA_GPU device"
]
[PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU')]