'ExecutionEnvironment',"multi-gpu"
08.05.2020
Коллеги, добрый день.
Почему может не работать тренеровка сети на 2-х GPU.
gpuDevice(1)
ans =
CUDADevice with properties:
Name: 'Quadro P2000'
Index: 1
ComputeCapability: '6.1'
SupportsDouble: 1
DriverVersion: 10.2000
ToolkitVersion: 10.1000
MaxThreadsPerBlock: 1024
MaxShmemPerBlock: 49152
MaxThreadBlockSize: [1024 1024 64]
MaxGridSize: [2.1475e+09 65535 65535]
SIMDWidth: 32
TotalMemory: 5.3687e+09
AvailableMemory: 4.3120e+09
MultiprocessorCount: 8
ClockRateKHz: 1480500
ComputeMode: 'Default'
GPUOverlapsTransfers: 1
KernelExecutionTimeout: 1
CanMapHostMemory: 1
DeviceSupported: 1
DeviceSelected: 1
gpuDevice(2)
ans =
CUDADevice with properties:
Name: 'Quadro P2000'
Index: 2
ComputeCapability: '6.1'
SupportsDouble: 1
DriverVersion: 10.2000
ToolkitVersion: 10.1000
MaxThreadsPerBlock: 1024
MaxShmemPerBlock: 49152
MaxThreadBlockSize: [1024 1024 64]
MaxGridSize: [2.1475e+09 65535 65535]
SIMDWidth: 32
TotalMemory: 5.3687e+09
AvailableMemory: 4.3363e+09
MultiprocessorCount: 8
ClockRateKHz: 1480500
ComputeMode: 'Default'
GPUOverlapsTransfers: 1
KernelExecutionTimeout: 1
CanMapHostMemory: 1
DeviceSupported: 1
DeviceSelected: 1
Ответы