AWS Amazon EC2

g6.8xlarge

⚡ 1x GPU • 24 GB

Part of the G6 Accelerated Computing family — Hardware accelerators (GPUs, FPGAs) for specialized compute.

Name:gGPU Graphics & ML Inference66th Generation8xlarge8 Extra Large

On-demand, Linux

$2.0144/ hr

~$1,470.51/mo · lowest in US East (N. Virginia), us-east-1

Savings Plan, Linux

$1.15576/ hr

From, with a 3-year Compute Savings Plan, No Upfront: −43% vs on-demand rate (us-east-1)

Per unit

$0.0630/ vCPU-hr

$0.0157/GB-hr

Compute

32vCPUs

16 cores • x86_64

Memory

128GB

4.0 GB / vCPU

Network

25 Gbps

Baseline: 25 Gbps

Storage & EBS

2,000MB/s

Max: 64,000 IOPS

Alternatives to g6.8xlarge

Option	Price vs g6.8xlarge	Why look at it
g5.8xlarge Compare →	+22%+$0.434/hr Linux on-demand, us-east-1	Previous generation, same vCPU and RAM
g4dn.8xlarge Compare →	+8%+$0.162/hr Linux on-demand, us-east-1	Intel (x86_64), older generation, same vCPU and RAM
inf2.8xlarge Compare →	-2%-$0.0465/hr Linux on-demand, us-east-1	Older generation, same vCPU and RAM

On Google Cloud

g6.8xlarge 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 / Threads32 vCPUs

Physical Cores16

Clock Speed2.6 GHz

Intel AVX / AVX2No

Intel Turbo BoostNo

Bare MetalNo (Virtualized)

Performance & Benchmarks

CoreMark Score689,915

CoreMark / vCPU21,559.9 / core

FFmpeg Transcoding135 FPS

CoreMark per Dollar342,492 score / $

Cost per vCPU-hr$0.0630

Cost per GB-hr$0.0157

Containers & Kubernetes

EKS Max Pods234 pods

Max ENIs8

IPv4/v6 per ENI30

Total Addressable IPs240 total IPs

CNI Prefix DelegationSupported (High Density)

Networking & Bandwidth

Network Performance25 Gbps

Baseline Bandwidth25 Gbps

Burst / Peak Bandwidth25 Gbps

Enhanced NetworkingYes

EFA SupportNo

IPv6 SupportNative Dual-Stack Supported

Storage & EBS Optimization

Storage ConfigurationNVMe SSD (900 GB)

Local NVMe Disk900 GB NVMe SSD

EBS Baseline Throughput2000 MB/s

EBS Max Throughput2000 MB/s

EBS Baseline IOPS64,000

EBS Max / Burst IOPS64,000

Dedicated EBS Throughput16000 Mbps

GPU & Hardware Acceleration

GPU ModelNVIDIA L4

Accelerator Count1x GPU(s)

Total GPU Memory24.0 GB

Memory Architecture—

Acceleration RuntimesCUDA, TensorRT, cuDNN, PyTorch, PyJAX

Platform & Architecture Details

Hypervisor / SystemAWS Nitro System (PCIe Offload)

Nitro EnclavesSupported (Isolated security enclaves)

Normalization Factor2.502982107

Generation StatusCurrent Generation (Recommended)

Rates by region

Live on-demand rates across all global AWS cloud regions

Lowest Price Region

$2.0144/ hr

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

Highest Price Region

$3.4241/ hr

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

Global Fleet Average

$2.4628/ hr

Across all active regions

Regional Coverage

20regions

Available worldwide

Linux — Available in all 20 regions
Region	On-Demand Price	% vs On-Demand Rate
ap-south-2 Asia Pacific (Hyderabad)	$2.4189
ap-southeast-5 Asia Pacific (Malaysia)	$2.5375
ap-south-1 Asia Pacific (Mumbai)	$2.4189
ap-northeast-2 Asia Pacific (Seoul)	$2.4769
ap-southeast-2 Asia Pacific (Sydney)	$2.6191
ap-northeast-1 Asia Pacific (Tokyo)	$2.9215
us-gov-east-1 AWS GovCloud (US-East)	$2.5390
us-gov-west-1 AWS GovCloud (US-West)	$2.5390
ca-central-1 Canada (Central)	$2.2367
eu-central-1 EU (Frankfurt)	$2.5190
eu-west-2 EU (London)	$2.5570
eu-west-3 EU (Paris)	$2.5570
eu-north-1 EU (Stockholm)	$2.1366
eu-south-2 Europe (Spain)	$2.1225
eu-central-2 Europe (Zurich)	$2.7157
me-central-1 Middle East (UAE)	$2.4735
sa-east-1 South America (Sao Paulo) Highest	$3.4241
us-east-1 US East (N. Virginia) Lowest	$2.0144
us-east-2 US East (Ohio) Lowest	$2.0144
us-west-2 US West (Oregon) Lowest	$2.0144

Windows — Available in all 20 regions
Region	On-Demand Price	% vs On-Demand Rate
ap-south-2 Asia Pacific (Hyderabad)	$3.8909
ap-southeast-5 Asia Pacific (Malaysia)	$4.0095
ap-south-1 Asia Pacific (Mumbai)	$3.8909
ap-northeast-2 Asia Pacific (Seoul)	$3.9489
ap-southeast-2 Asia Pacific (Sydney)	$4.0911
ap-northeast-1 Asia Pacific (Tokyo)	$4.3935
us-gov-east-1 AWS GovCloud (US-East)	$4.0110
us-gov-west-1 AWS GovCloud (US-West)	$4.0110
ca-central-1 Canada (Central)	$3.7087
eu-central-1 EU (Frankfurt)	$3.9910
eu-west-2 EU (London)	$4.0290
eu-west-3 EU (Paris)	$4.0290
eu-north-1 EU (Stockholm)	$3.6086
eu-south-2 Europe (Spain)	$3.5945
eu-central-2 Europe (Zurich)	$4.1877
me-central-1 Middle East (UAE)	$3.9455
sa-east-1 South America (Sao Paulo) Highest	$4.8961
us-east-1 US East (N. Virginia) Lowest	$3.4864
us-east-2 US East (Ohio) Lowest	$3.4864
us-west-2 US West (Oregon) Lowest	$3.4864

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 g6.8xlarge on Linux is 57% of the on-demand rate, so the instance has to run at least 57% 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.