AWS Amazon EC2

p5e.48xlarge

⚡ 8x GPU • 1128.0 GB HBM3

Part of the P5E Accelerated Computing family — Hardware accelerators (GPUs, FPGAs) for specialized compute.

Name:pGPU Deep Learning & HPC55th GenerationeExtra Capacity / Expansion48xlarge48 Extra Large

On-demand, Linux

$63.2960/ hr

~$46,206.08/mo · lowest in US East (Ohio), us-east-2

Savings Plan, Linux

$46.52491/ hr

From, with a 3-year Compute Savings Plan, No Upfront: −26% vs on-demand rate (us-east-2)

Per unit

$0.3297/ vCPU-hr

$0.0309/GB-hr

Compute

192vCPUs

96 cores • x86_64

Memory

2,048GB

10.7 GB / vCPU

Network

3200 Gbps

Baseline: 3200 Gbps

Storage & EBS

10,000MB/s

Max: 260,000 IOPS

On Google Cloud

p5e.48xlarge is an accelerated machine. Those are compared by the accelerator, not by vCPU and memory, so this page doesn't pick a match. Browse Google Cloud machines.

Technical Specifications & Limits

Hardware specs, container thresholds, and system limits

Compute & Processor

ProcessorAMD EPYC 7R13 Processor

Architecturex86_64 (x86_64)

vCPUs / Threads192 vCPUs

Physical Cores96

Clock Speed2.95 GHz

Intel AVX / AVX2No

Intel Turbo BoostNo

Bare MetalNo (Virtualized)

Performance & Benchmarks

CoreMark Score3,884,477

CoreMark / vCPU20,231.7 / core

FFmpeg Transcoding—

CoreMark per Dollar61,370 score / $

Cost per vCPU-hr$0.3297

Cost per GB-hr$0.0309

Containers & Kubernetes

EKS Max Pods100 pods

Max ENIs64

IPv4/v6 per ENI50

Total Addressable IPs3200 total IPs

CNI Prefix DelegationSupported (High Density)

Networking & Bandwidth

Network Performance3200 Gbps

Baseline Bandwidth3200 Gbps

Burst / Peak Bandwidth3200 Gbps

Enhanced NetworkingYes

EFA SupportNo

IPv6 SupportNative Dual-Stack Supported

Storage & EBS Optimization

Storage ConfigurationSSD (30720 GB)

Local NVMe Disk30,720 GB SSD

EBS Baseline Throughput10000 MB/s

EBS Max Throughput10000 MB/s

EBS Baseline IOPS260,000

EBS Max / Burst IOPS260,000

Dedicated EBS Throughput80 Gbps

GPU & Hardware Acceleration

GPU ModelNVIDIA H200

Accelerator Count8x GPU(s)

Total GPU Memory1128.0 GB

Memory ArchitectureHBM3

Acceleration RuntimesCUDA, TensorRT, cuDNN, PyTorch, PyJAX

Platform & Architecture Details

Hypervisor / SystemAWS Nitro System (PCIe Offload)

Nitro EnclavesSupported (Isolated security enclaves)

Normalization Factor384

Generation StatusCurrent Generation (Recommended)

Rates by region

Live on-demand rates across all global AWS cloud regions

Lowest Price Region

$63.2960/ hr

US East (Ohio) (us-east-2)

Highest Price Region

$82.2848/ hr

Asia Pacific (Sydney) (ap-southeast-2)

Global Fleet Average

$72.7904/ hr

Across all active regions

Regional Coverage

4regions

Available worldwide

Linux — Available in all 4 regions
Region	On-Demand Price	% vs On-Demand Rate
ap-southeast-2 Asia Pacific (Sydney) Highest	$82.2848
eu-west-2 EU (London) Highest	$82.2848
us-east-2 US East (Ohio) Lowest	$63.2960
us-west-2 US West (Oregon) Lowest	$63.2960

Savings Plans: what to know

A Savings Plan commits you to a dollar amount per hour of spend for one or three years, not to an instance: usage up to that amount is billed at the plan's rates and the rest at on-demand, and unused commitment is still billed.
A Compute Savings Plan follows your usage across instance families, sizes, regions and operating systems; an EC2 Instance Savings Plan is locked to one family in one region in exchange for a lower rate (what that flexibility costs).
In US East (Ohio) (us-east-2), the 3-year Compute Savings Plan rate (No Upfront) for p5e.48xlarge on Linux is 74% of the on-demand rate, so the instance has to run at least 74% of the hours for the commitment to cost less than on-demand.
Rates are AWS's effective hourly rate for each payment option; All Upfront pays the whole term's commitment at purchase, which is why its hourly rate is lower; Partial Upfront (at least half up front) isn't shown on this page. The Savings Plans calculator has all three and sizes a commitment for this instance or a mix of them. How AWS applies a plan each hour: how Compute Savings Plans decide what they cover. Sizing one for elastic workloads: managing Compute Savings Plans.