AWS Amazon EC2

p5.4xlarge

⚡ 1x GPU • 80.0 GB HBM3

Part of the P5 Accelerated Computing family — Hardware accelerators (GPUs, FPGAs) for specialized compute.

Name:pGPU Deep Learning & HPC55th Generation4xlarge4 Extra Large

On-demand, Linux

$6.8800/ hr

~$5,022.40/mo · lowest in US East (N. Virginia), us-east-1

Savings Plan, Linux

$5.05267/ hr

From, with a 3-year Compute Savings Plan, No Upfront: −27% vs on-demand rate (us-east-1)

Per unit

$0.4300/ vCPU-hr

$0.0269/GB-hr

Compute

16vCPUs

8 cores • x86_64

Memory

256GB

16.0 GB / vCPU

Network

100 Gbps

Baseline: 100 Gbps

Storage & EBS

1,250MB/s

Max: 32,500 IOPS

On Google Cloud

p5.4xlarge 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 / Threads16 vCPUs

Physical Cores8

Clock Speed2.95 GHz

Intel AVX / AVX2No

Intel Turbo BoostNo

Bare MetalNo (Virtualized)

Performance & Benchmarks

CoreMark ScoreNot benchmarked

CoreMark / vCPU—

FFmpeg Transcoding—

CoreMark per Dollar—

Cost per vCPU-hr$0.4300

Cost per GB-hr$0.0269

Containers & Kubernetes

EKS Max Pods118 pods

Max ENIs4

IPv4/v6 per ENI30

Total Addressable IPs120 total IPs

CNI Prefix DelegationSupported (High Density)

Networking & Bandwidth

Network Performance100 Gbps

Baseline Bandwidth100 Gbps

Burst / Peak Bandwidth100 Gbps

Enhanced NetworkingYes

EFA SupportNo

IPv6 SupportNative Dual-Stack Supported

Storage & EBS Optimization

Storage ConfigurationSSD (3840 GB)

Local NVMe Disk3,840 GB SSD

EBS Baseline Throughput1250 MB/s

EBS Max Throughput1250 MB/s

EBS Baseline IOPS32,500

EBS Max / Burst IOPS32,500

Dedicated EBS Throughput10 Gbps

GPU & Hardware Acceleration

GPU ModelNVIDIA H100

Accelerator Count1x GPU(s)

Total GPU Memory80.0 GB

Memory ArchitectureHBM3

Acceleration RuntimesCUDA, TensorRT, cuDNN, PyTorch, PyJAX

Platform & Architecture Details

Hypervisor / SystemAWS Nitro System (PCIe Offload)

Nitro EnclavesSupported (Isolated security enclaves)

Normalization Factor48

Generation StatusCurrent Generation (Recommended)

Rates by region

Live on-demand rates across all global AWS cloud regions

Lowest Price Region

$6.8800/ hr

US East (N. Virginia) (us-east-1)

Highest Price Region

$11.5584/ hr

South America (Sao Paulo) (sa-east-1)

Global Fleet Average

$8.3879/ hr

Across all active regions

Regional Coverage

12regions

Available worldwide

Linux — Available in all 12 regions
Region	On-Demand Price	% vs On-Demand Rate
ap-southeast-3 Asia Pacific (Jakarta)	$9.4944
ap-south-1 Asia Pacific (Mumbai)	$8.2560
ap-southeast-2 Asia Pacific (Sydney)	$8.9440
ap-northeast-1 Asia Pacific (Tokyo)	$8.6000
us-gov-west-1 AWS GovCloud (US-West)	$8.2560
eu-central-1 EU (Frankfurt)	$8.6000
eu-west-2 EU (London)	$8.9440
eu-north-1 EU (Stockholm)	$7.3616
sa-east-1 South America (Sao Paulo) Highest	$11.5584
us-east-1 US East (N. Virginia) Lowest	$6.8800
us-east-2 US East (Ohio) Lowest	$6.8800
us-west-2 US West (Oregon) Lowest	$6.8800

Windows — Available in all 12 regions
Region	On-Demand Price	% vs On-Demand Rate
ap-southeast-3 Asia Pacific (Jakarta)	$10.2304
ap-south-1 Asia Pacific (Mumbai)	$8.9920
ap-southeast-2 Asia Pacific (Sydney)	$9.6800
ap-northeast-1 Asia Pacific (Tokyo)	$9.3360
us-gov-west-1 AWS GovCloud (US-West)	$8.9920
eu-central-1 EU (Frankfurt)	$9.3360
eu-west-2 EU (London)	$9.6800
eu-north-1 EU (Stockholm)	$8.0976
sa-east-1 South America (Sao Paulo) Highest	$12.2944
us-east-1 US East (N. Virginia) Lowest	$7.6160
us-east-2 US East (Ohio) Lowest	$7.6160
us-west-2 US West (Oregon) Lowest	$7.6160

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 (N. Virginia) (us-east-1), the 3-year Compute Savings Plan rate (No Upfront) for p5.4xlarge on Linux is 73% of the on-demand rate, so the instance has to run at least 73% of the hours for the commitment to cost less than on-demand.
On Windows a plan discounts only the compute part of the rate: the Windows license part costs the same as on demand, so the percentage off is smaller than on Linux.
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.