AWS Amazon EC2

g5.48xlarge

⚡ 8x GPU • 192 GB

Part of the G5 Accelerated Computing family — Hardware accelerators (GPUs, FPGAs) for specialized compute.

Name:gGPU Graphics & ML Inference55th Generation48xlarge48 Extra Large

On-demand, Linux

$16.2880/ hr

~$11,890.24/mo · lowest in US East (N. Virginia), us-east-1

Savings Plan, Linux

$8.79552/ hr

From, with a 3-year Compute Savings Plan, No Upfront: −46% vs on-demand rate (us-east-1)

Per unit

$0.0848/ vCPU-hr

$0.0212/GB-hr

Compute

192vCPUs

96 cores • x86_64

Memory

768GB

4.0 GB / vCPU

Network

100 Gbps

Baseline: 100 Gbps

Storage & EBS

2,375MB/s

Max: 80,000 IOPS

Alternatives to g5.48xlarge

Option	Price vs g5.48xlarge	Why look at it
g6.48xlarge Compare →	-18%-$2.94/hr Linux on-demand, us-east-1	Newer generation, same vCPU and RAM
inf2.48xlarge Compare →	-20%-$3.31/hr Linux on-demand, us-east-1	Older generation, same vCPU and RAM

On Google Cloud

g5.48xlarge 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 7R32

Architecturex86_64 (x86_64)

vCPUs / Threads192 vCPUs

Physical Cores96

Clock Speed2.8 GHz

Intel AVX / AVX2No

Intel Turbo BoostNo

Bare MetalNo (Virtualized)

Performance & Benchmarks

CoreMark Score3,614,458

CoreMark / vCPU18,825.3 / core

FFmpeg Transcoding110 FPS

CoreMark per Dollar221,909 score / $

Cost per vCPU-hr$0.0848

Cost per GB-hr$0.0212

Containers & Kubernetes

EKS Max Pods345 pods

Max ENIs7

IPv4/v6 per ENI50

Total Addressable IPs350 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 ConfigurationNVMe SSD (7600 GB)

Local NVMe Disk7,600 GB NVMe SSD

EBS Baseline Throughput2375 MB/s

EBS Max Throughput2375 MB/s

EBS Baseline IOPS80,000

EBS Max / Burst IOPS80,000

Dedicated EBS Throughput19000 Mbps

GPU & Hardware Acceleration

GPU ModelNVIDIA A10G

Accelerator Count8x GPU(s)

Total GPU Memory192.0 GB

Memory Architecture—

Acceleration RuntimesCUDA, TensorRT, cuDNN, PyTorch, PyJAX

Platform & Architecture Details

Hypervisor / SystemAWS Nitro System (PCIe Offload)

Nitro EnclavesSupported (Isolated security enclaves)

Normalization Factor16.190854871

Generation StatusCurrent Generation (Recommended)

Rates by region

Live on-demand rates across all global AWS cloud regions

Lowest Price Region

$16.2880/ hr

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

Highest Price Region

$27.6864/ hr

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

Global Fleet Average

$20.0161/ hr

Across all active regions

Regional Coverage

19regions

Available worldwide

Linux — Available in all 19 regions
Region	On-Demand Price	% vs On-Demand Rate
ap-east-1 Asia Pacific (Hong Kong)	$25.0764
ap-southeast-3 Asia Pacific (Jakarta)	$22.7967
ap-south-1 Asia Pacific (Mumbai)	$19.5586
ap-northeast-2 Asia Pacific (Seoul)	$20.0281
ap-southeast-2 Asia Pacific (Sydney)	$21.1776
ap-northeast-1 Asia Pacific (Tokyo)	$23.6225
ca-central-1 Canada (Central)	$18.0852
eu-central-1 EU (Frankfurt)	$20.3681
eu-west-1 EU (Ireland)	$18.1823
eu-west-2 EU (London)	$20.6757
eu-west-3 EU (Paris)	$20.6757
eu-north-1 EU (Stockholm)	$17.2756
eu-south-2 Europe (Spain)	$17.1623
il-central-1 Israel (Tel Aviv)	$19.0914
me-central-1 Middle East (UAE)	$19.9795
sa-east-1 South America (Sao Paulo) Highest	$27.6864
us-east-1 US East (N. Virginia) Lowest	$16.2880
us-east-2 US East (Ohio) Lowest	$16.2880
us-west-2 US West (Oregon) Lowest	$16.2880

Windows — Available in all 19 regions
Region	On-Demand Price	% vs On-Demand Rate
ap-east-1 Asia Pacific (Hong Kong)	$33.9084
ap-southeast-3 Asia Pacific (Jakarta)	$31.6287
ap-south-1 Asia Pacific (Mumbai)	$28.3906
ap-northeast-2 Asia Pacific (Seoul)	$28.8601
ap-southeast-2 Asia Pacific (Sydney)	$30.0096
ap-northeast-1 Asia Pacific (Tokyo)	$32.4545
ca-central-1 Canada (Central)	$26.9172
eu-central-1 EU (Frankfurt)	$29.2001
eu-west-1 EU (Ireland)	$27.0143
eu-west-2 EU (London)	$29.5077
eu-west-3 EU (Paris)	$29.5077
eu-north-1 EU (Stockholm)	$26.1076
eu-south-2 Europe (Spain)	$25.9943
il-central-1 Israel (Tel Aviv)	$27.9234
me-central-1 Middle East (UAE)	$28.8115
sa-east-1 South America (Sao Paulo) Highest	$36.5184
us-east-1 US East (N. Virginia) Lowest	$25.1200
us-east-2 US East (Ohio) Lowest	$25.1200
us-west-2 US West (Oregon) Lowest	$25.1200

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 g5.48xlarge on Linux is 54% of the on-demand rate, so the instance has to run at least 54% 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.