Class: Google::Cloud::Container::V1::GPUSharingConfig

Inherits:

Object

Object
Google::Cloud::Container::V1::GPUSharingConfig

show all

Extended by:: Protobuf::MessageExts::ClassMethods

Includes:: Protobuf::MessageExts

Defined in:: proto_docs/google/container/v1/cluster_service.rb

Overview

GPUSharingConfig represents the GPU sharing configuration for Hardware Accelerators.

Defined Under Namespace

Modules: GPUSharingStrategy

Instance Attribute Summary collapse

#gpu_sharing_strategy ⇒ ::Google::Cloud::Container::V1::GPUSharingConfig::GPUSharingStrategy
The type of GPU sharing strategy to enable on the GPU node.
#max_shared_clients_per_gpu ⇒ ::Integer
The max number of containers that can share a physical GPU.

Instance Attribute Details

#gpu_sharing_strategy ⇒ `::Google::Cloud::Container::V1::GPUSharingConfig::GPUSharingStrategy`

Returns The type of GPU sharing strategy to enable on the GPU node.

Returns:

(::Google::Cloud::Container::V1::GPUSharingConfig::GPUSharingStrategy) —
The type of GPU sharing strategy to enable on the GPU node.

# File 'proto_docs/google/container/v1/cluster_service.rb', line 6106

class GPUSharingConfig
  include ::Google::Protobuf::MessageExts
  extend ::Google::Protobuf::MessageExts::ClassMethods

  # The type of GPU sharing strategy currently provided.
  module GPUSharingStrategy
    # Default value.
    GPU_SHARING_STRATEGY_UNSPECIFIED = 0

    # GPUs are time-shared between containers.
    TIME_SHARING = 1

    # GPUs are shared between containers with NVIDIA MPS.
    MPS = 2
  end
end

#max_shared_clients_per_gpu ⇒ `::Integer`

Returns The max number of containers that can share a physical GPU.