Enum AcceleratorType

Hardware accelerator categories available for EC2 instances.

Namespace: Amazon.CDK.AWS.EC2

Assembly: Amazon.CDK.Lib.dll

Syntax (csharp)

public enum AcceleratorType

Syntax (vb)

Public Enum AcceleratorType

Remarks

Defines the general type of hardware accelerator that can be attached to an instance, typically used in instance requirement specifications (e.g., GPUs for compute-intensive tasks, FPGAs for custom logic, or inference chips for ML workloads).

ExampleMetadata: infused

Examples

Role infrastructureRole;
             InstanceProfile instanceProfile;
             Vpc vpc;


             var miCapacityProvider = new ManagedInstancesCapacityProvider(this, "MICapacityProvider", new ManagedInstancesCapacityProviderProps {
                 InfrastructureRole = infrastructureRole,
                 Ec2InstanceProfile = instanceProfile,
                 Subnets = vpc.PrivateSubnets,
                 InstanceRequirements = new InstanceRequirementsConfig {
                     // Required: CPU and memory constraints
                     VCpuCountMin = 2,
                     VCpuCountMax = 8,
                     MemoryMin = Size.Gibibytes(4),
                     MemoryMax = Size.Gibibytes(32),

                     // CPU preferences
                     CpuManufacturers = new [] { CpuManufacturer.INTEL, CpuManufacturer.AMD },
                     InstanceGenerations = new [] { InstanceGeneration.CURRENT },

                     // Instance type filtering
                     AllowedInstanceTypes = new [] { "m5.*", "c5.*" },

                     // Performance characteristics
                     BurstablePerformance = BurstablePerformance.EXCLUDED,
                     BareMetal = BareMetal.EXCLUDED,

                     // Accelerator requirements (for ML/AI workloads)
                     AcceleratorTypes = new [] { AcceleratorType.GPU },
                     AcceleratorManufacturers = new [] { AcceleratorManufacturer.NVIDIA },
                     AcceleratorNames = new [] { AcceleratorName.T4, AcceleratorName.V100 },
                     AcceleratorCountMin = 1,

                     // Storage requirements
                     LocalStorage = LocalStorage.REQUIRED,
                     LocalStorageTypes = new [] { LocalStorageType.SSD },
                     TotalLocalStorageGBMin = 100,

                     // Network requirements
                     NetworkInterfaceCountMin = 2,
                     NetworkBandwidthGbpsMin = 10,

                     // Cost optimization
                     OnDemandMaxPricePercentageOverLowestPrice = 10
                 }
             });

Synopsis

Fields

FPGA	Field Programmable Gate Array accelerators, such as Xilinx FPGAs.
GPU	Graphics Processing Unit accelerators, such as NVIDIA GPUs.
INFERENCE	Inference accelerators, such as AWS Inferentia.

Fields

Name	Description
FPGA	Field Programmable Gate Array accelerators, such as Xilinx FPGAs.
GPU	Graphics Processing Unit accelerators, such as NVIDIA GPUs.
INFERENCE	Inference accelerators, such as AWS Inferentia.