ml.p3.8xlarge

The ml.p3.8xlarge is a aws sagemaker instance type featuring multiple vCPUs and optimized memory, commonly used for sagemaker workloads.

Last updated at Fri May 08 2026
Instance Info
Compute Value
vCPUs 32
Memory 244 GiB
Physical Processor Intel Xeon E5-2686 v4 (Broadwell)
Clock Speed 2.3 GHz
CPU Architecture 64-bit
GPU 4
GPU Memory 64
Network Value
Network Performance 10 Gigabit
Storage Value
Storage EBS only
Amazon Value
Current Generation Yes
Instance Type ml.p3.8xlarge
Family Instances
Instance vCPUs Memory
ml.p3.2xlarge 8 61
ml.p3.8xlarge 32 244
ml.p3.16xlarge 64 488