Configure the hybrid job instance to run your algorithm script - Amazon Braket

Configure the hybrid job instance to run your algorithm script

Depending on your algorithm, you may have different requirements. By default, Amazon Braket runs your algorithm script on an ml.m5.large instance. However, you can customize this instance type when you create a hybrid job using the following import and configuration argument.

from braket.jobs.config import InstanceConfig

job = AwsQuantumJob.create(
    ...
    instance_config=InstanceConfig(instanceType="ml.p3.8xlarge"), # Use NVIDIA Tesla V100 instance with 4 GPUs.
    ...
    ),

If you are running an embedded simulation and have specified a local device in the device configuration, you will be able to additionally request more than one instance in the InstanceConfig by specifying the instanceCount and setting it to be greater than one. The upper limit is 5. For instance, you can choose 3 instances as follows.

from braket.jobs.config import InstanceConfig
job = AwsQuantumJob.create(
    ...
    instance_config=InstanceConfig(instanceType="ml.p3.8xlarge", instanceCount=3), # Use 3 NVIDIA Tesla V100
    ...
    ),

When you use multiple instances, consider distributing your hybrid job using the data parallel feature. See the following example notebook for more details on how-to see this Braket example.

The following three tables list the available instance types and specs for standard, compute optimized, and accelerated computing instances.

Note

To view the default classical compute instance quotas for Hybrid Jobs, see this page.

Standard Instances vCPU Memory

ml.m5.large (default)

2

8 GiB

ml.m5.xlarge

4

16 GiB

ml.m5.2xlarge

8

32 GiB

ml.m5.4xlarge

16

64 GiB

ml.m5.12xlarge

48

192 GiB

ml.m5.24xlarge

96

384 GiB

ml.m4.xlarge

4

16 GiB

ml.m4.2xlarge

8

32 GiB

ml.m4.4xlarge

16

64 GiB

ml.m4.10xlarge

40

256 GiB
Compute Optimized Instances vCPU Memory

ml.c4.xlarge

4

7.5 GiB

ml.c4.2xlarge

8

15 GiB

ml.c4.4xlarge

16

30 GiB

ml.c4.8xlarge

36

192 GiB

ml.c5.xlarge

4

8 GiB

ml.c5.2xlarge

8

16 GiB

ml.c5.4xlarge

16

32 GiB

ml.c5.9xlarge

36

72 GiB

ml.c5.18xlarge

72

144 GiB

ml.c5n.xlarge

4

10.5 GiB

ml.c5n.2xlarge

8

21 GiB

ml.c5n.4xlarge

16

42 GiB

ml.c5n.9xlarge

36

96 GiB

ml.c5n.18xlarge

72

192 GiB
Accelerated Compting Instances vCPU Memory

ml.p2.xlarge

4

61 GiB

ml.p2.8xlarge

32

488 GiB

ml.p2.16xlarge

64

732 GiB

ml.p3.2xlarge

8

61 GiB

ml.p3.8xlarge

32

244 GiB

ml.p3.16xlarge

64

488 GiB

ml.g4dn.xlarge

4

16 GiB

ml.g4dn.2xlarge

8

32 GiB

ml.g4dn.4xlarge

16

64 GiB

ml.g4dn.8xlarge

32

128 GiB

ml.g4dn.12xlarge

48

192 GiB

ml.g4dn.16xlarge

64

256 GiB
Note

p3 instances are not available in us-west-1. If your hybrid job is unable to provision requested ML compute capacity, use another Region.

Each instance uses a default configuration of data storage (SSD) of 30 GB. But you can adjust the storage in the same way that you configure the instanceType. The following example shows how to increase the total storage to 50 GB.

from braket.jobs.config import InstanceConfig

job = AwsQuantumJob.create(
    ...
    instance_config=InstanceConfig(
        instanceType="ml.p3.8xlarge",
        volumeSizeInGb=50,
    ),
    ...
    ),