← All instances

g5.4xlarge

⚡ 1x GPU • 24.0 GB

Part of the G5 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.55th Generation5 = 5th GenerationAWS 5th generation hardware platform with updated CPU/system architecture.4xlarge4 Extra Large4xlarge = 4 Extra Large16 vCPUs • 64.0 GiB RAM • 600 GB NVMe SSD • Up to 25 Gigabit Network • 1x NVIDIA A10G (24.0 GB VRAM)
Price (Linux)
$1.6240/ hr
~$1,185.52/mo
Compute
16vCPUs
8 cores • x86_64
Memory
64.0GiB
4.0 GiB / vCPU
Value
$0.1015/ vCPU-hr
$0.0254/GiB-hr
Network
25Gbps
Baseline: 10 Gbps
Storage & EBS
593.75MB/s
Max: 20,000 IOPS

Technical Specifications & Limits

Hardware specs, container thresholds, and system limits

Compute & Processor

ProcessorAMD EPYC 7R32
Architecturex86_64 (x86_64)
vCPUs / Threads16 vCPUs
Physical Cores8
Clock Speed2.8 GHz
Intel AVX / AVX2No
Intel Turbo BoostNo
Bare MetalNo (Virtualized)

Performance & Benchmarks

CoreMark Score312,480
CoreMark / vCPU19,530 / core
FFmpeg Transcoding111 FPS
CoreMark per Dollar192,414 score / $
Cost per vCPU-hr$0.1015
Cost per GiB-hr$0.0254

Containers & Kubernetes

EKS Max Pods234 pods
Max ECS Tasks40 tasks
Max ENIs8
IPv4/v6 per ENI30
Total Addressable IPs240 total IPs
CNI Prefix DelegationSupported (High Density)

Networking & Bandwidth

Network PerformanceUp to 25 Gigabit
Baseline Bandwidth10 Gbps
Burst / Peak Bandwidth25 Gbps
Enhanced NetworkingYes
EFA SupportNo
IPv6 SupportNative Dual-Stack Supported

Storage & EBS Optimization

Storage ConfigurationNVMe SSD (600 GB)
Local NVMe Disk600 GB
EBS Baseline Throughput593.75 MB/s
EBS Max Throughput593.75 MB/s
EBS Baseline IOPS20,000
EBS Max / Burst IOPS20,000
Dedicated EBS Throughput8000 Mbps

GPU & Hardware Acceleration

GPU ModelNVIDIA A10G
Accelerator Count1x GPU(s)
Total GPU Memory24.0 GB
Memory ArchitectureGDDR / HBM
CUDA Compute Capabilityv8.6
Acceleration RuntimesCUDA, TensorRT, cuDNN, PyTorch, PyJAX

Platform & Architecture Details

Hypervisor / SystemAWS Nitro System (PCIe Offload)
Nitro EnclavesSupported (Isolated security enclaves)
Normalization Factor1.614314115
Generation StatusCurrent Generation (Recommended)
Date Introduced2021-11-12

Regional Pricing Matrix

Live on-demand rates across all global AWS cloud regions
Lowest Price Region
$1.6240/ hr
US East (N. Virginia) (us-east-1)
Highest Price Region
$2.7605/ hr
South America (Sao Paulo) (sa-east-1)
Global Fleet Average
$2.0075/ hr
Across all active regions
Regional Coverage
20regions
Available worldwide
Linux — Available in all 20 regions
Region NameRegion CodeOn-Demand Price
Asia Pacific (Hong Kong)
ap-east-1$2.5002
Asia Pacific (Jakarta)
ap-southeast-3$2.2729
Asia Pacific (Mumbai)
ap-south-1$1.9501
Asia Pacific (Seoul)
ap-northeast-2$1.9969
Asia Pacific (Sydney)
ap-southeast-2$2.1115
Asia Pacific (Tokyo)
ap-northeast-1$2.3553
Canada (Central)
ca-central-1$1.8032
EU (Frankfurt)
eu-central-1$2.0308
EU (Ireland)
eu-west-1$1.8129
EU (London)
eu-west-2$2.0615
EU (Paris)
eu-west-3$2.0615
EU (Stockholm)
eu-north-1$1.7225
Europe (Spain)
eu-south-2$1.7112
Israel (Tel Aviv)
il-central-1$1.9035
Middle East (UAE)
me-central-1$1.9921
South America (Sao Paulo)Highest
sa-east-1$2.7605
US East (N. Virginia)Lowest
us-east-1$1.6240
US East (Ohio)Lowest
us-east-2$1.6240
US West (Los Angeles)
us-west-2-lax-1$2.2310
US West (Oregon)Lowest
us-west-2$1.6240