← All instances

g6.12xlarge

⚡ 4x GPU • 96.0 GB

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

Name:gGPU Graphics & ML Inferenceg = GPU Graphics & ML InferencePowered by GPUs (NVIDIA/AMD) for graphics-intensive applications, video encoding, remote workstations, and machine learning inference.66th Generation6 = 6th GenerationAWS 6th generation hardware platform with updated CPU/system architecture.12xlarge12 Extra Large12xlarge = 12 Extra Large48 vCPUs • 192.0 GiB RAM • 4 GB X 940 GB NVMe SSD • 40 Gigabit Network • 4x NVIDIA L4 (96.0 GB VRAM)
Price (Linux)
$4.6016/ hr
~$3,359.17/mo
Compute
48vCPUs
24 cores • x86_64
Memory
192.0GiB
4.0 GiB / vCPU
Value
$0.0959/ vCPU-hr
$0.0240/GiB-hr
Network
40Gbps
Baseline: 40 Gbps
Storage & EBS
2,500MB/s
Max: 80,000 IOPS

Technical Specifications & Limits

Hardware specs, container thresholds, and system limits

Compute & Processor

ProcessorAMD EPYC 7R13 Processor
Architecturex86_64 (x86_64)
vCPUs / Threads48 vCPUs
Physical Cores24
Clock Speed2.6 GHz
Intel AVX / AVX2No
Intel Turbo BoostNo
Bare MetalNo (Virtualized)

Performance & Benchmarks

CoreMark Score1,056,861
CoreMark / vCPU22,017.9 / core
FFmpeg Transcoding138 FPS
CoreMark per Dollar229,673 score / $
Cost per vCPU-hr$0.0959
Cost per GiB-hr$0.0240

Containers & Kubernetes

EKS Max Pods234 pods
Max ECS Tasks120 tasks
Max ENIs8
IPv4/v6 per ENI30
Total Addressable IPs240 total IPs
CNI Prefix DelegationSupported (High Density)

Networking & Bandwidth

Network Performance40 Gigabit
Baseline Bandwidth40 Gbps
Burst / Peak Bandwidth40 Gbps
Enhanced NetworkingYes
EFA SupportNo
IPv6 SupportNative Dual-Stack Supported

Storage & EBS Optimization

Storage ConfigurationX 940 GB NVMe SSD (4 GB)
Local NVMe Disk4 GB
EBS Baseline Throughput2500 MB/s
EBS Max Throughput2500 MB/s
EBS Baseline IOPS80,000
EBS Max / Burst IOPS80,000
Dedicated EBS Throughput20000 Mbps

GPU & Hardware Acceleration

GPU ModelNVIDIA L4
Accelerator Count4x GPU(s)
Total GPU Memory96.0 GB
Memory ArchitectureGDDR / HBM
CUDA Compute Capabilityv8.9
Acceleration RuntimesCUDA, TensorRT, cuDNN, PyTorch, PyJAX

Platform & Architecture Details

Hypervisor / SystemAWS Nitro System (PCIe Offload)
Nitro EnclavesSupported (Isolated security enclaves)
Normalization Factor5.717693837
Generation StatusCurrent Generation (Recommended)
Date Introduced2024-04-04

Regional Pricing Matrix

Live on-demand rates across all global AWS cloud regions
Lowest Price Region
$4.6016/ hr
US East (N. Virginia) (us-east-1)
Highest Price Region
$7.8218/ hr
South America (Sao Paulo) (sa-east-1)
Global Fleet Average
$5.6259/ hr
Across all active regions
Regional Coverage
20regions
Available worldwide
Linux — Available in all 20 regions
Region NameRegion CodeOn-Demand Price
Asia Pacific (Hyderabad)
ap-south-2$5.5256
Asia Pacific (Malaysia)
ap-southeast-5$5.7966
Asia Pacific (Mumbai)
ap-south-1$5.5256
Asia Pacific (Seoul)
ap-northeast-2$5.6582
Asia Pacific (Sydney)
ap-southeast-2$5.9830
Asia Pacific (Tokyo)
ap-northeast-1$6.6737
AWS GovCloud (US-East)
us-gov-east-1$5.8000
AWS GovCloud (US-West)
us-gov-west-1$5.8000
Canada (Central)
ca-central-1$5.1093
EU (Frankfurt)
eu-central-1$5.7543
EU (London)
eu-west-2$5.8412
EU (Paris)
eu-west-3$5.8412
EU (Stockholm)
eu-north-1$4.8806
Europe (Spain)
eu-south-2$4.8486
Europe (Zurich)
eu-central-2$6.2037
Middle East (UAE)
me-central-1$5.6505
South America (Sao Paulo)Highest
sa-east-1$7.8218
US East (N. Virginia)Lowest
us-east-1$4.6016
US East (Ohio)Lowest
us-east-2$4.6016
US West (Oregon)Lowest
us-west-2$4.6016