Compare instance

inf1.24xlarge

⚡ 16x GPU

Part of the INF1 Accelerated Computing family — Hardware accelerators (GPUs, FPGAs) for specialized compute.

Name:infAWS Inferentia Acceleratorsinf = AWS Inferentia AcceleratorsPurpose-built AWS Inferentia silicon designed for ultra-low latency, high-throughput machine learning inference.11st Generation1 = 1st GenerationAWS 1st generation hardware platform with updated CPU/system architecture.24xlarge24 Extra Large24xlarge = 24 Extra Large96 vCPUs • 192 GB RAM • EBS only • 100 Gbps Network • 16x GPU
Price (Linux)
$4.7210/ hr
~$3,446.33/mo
Compute
96vCPUs
48 cores • x86_64
Memory
192GB
2.0 GB / vCPU
Value
$0.0492/ vCPU-hr
$0.0246/GB-hr
Network
100 Gbps
Baseline: 100 Gbps
Storage & EBS
2,375MB/s
Max: 80,000 IOPS

Technical Specifications & Limits

Hardware specs, container thresholds, and system limits

Compute & Processor

ProcessorIntel Xeon Platinum 8275CL (Cascade Lake)
Architecturex86_64 (x86_64)
vCPUs / Threads96 vCPUs
Physical Cores48
Clock Speed—
Intel AVX / AVX2No
Intel Turbo BoostNo
Bare MetalNo (Virtualized)

Performance & Benchmarks

CoreMark Score1,539,880
CoreMark / vCPU16,040.4 / core
FFmpeg Transcoding194 FPS
CoreMark per Dollar326,177 score / $
Cost per vCPU-hr$0.0492
Cost per GB-hr$0.0246

Containers & Kubernetes

EKS Max Pods321 pods
Max ECS Tasks120 tasks
Max ENIs11
IPv4/v6 per ENI30
Total Addressable IPs330 total IPs
CNI Prefix DelegationSupported (High Density)

Networking & Bandwidth

Network Performance100 Gbps
Baseline Bandwidth100 Gbps
Burst / Peak Bandwidth100 Gbps
Enhanced NetworkingNo
EFA SupportNo
IPv6 SupportNative Dual-Stack Supported

Storage & EBS Optimization

Storage ConfigurationEBS only
Local NVMe DiskNone (EBS only)
EBS Baseline Throughput2375 MB/s
EBS Max Throughput2375 MB/s
EBS Baseline IOPS80,000
EBS Max / Burst IOPS80,000
Dedicated EBS Throughput14000 Mbps

GPU & Hardware Acceleration

GPU ModelNVIDIA Accelerator
Accelerator Count16x GPU(s)
Total GPU Memory—
Memory ArchitectureGDDR / HBM
CUDA Compute Capability—
Acceleration RuntimesCUDA, TensorRT, cuDNN, PyTorch, PyJAX

Platform & Architecture Details

Hypervisor / SystemAWS Nitro System (PCIe Offload)
Nitro EnclavesSupported (Isolated security enclaves)
Normalization FactorNA
Generation StatusCurrent Generation (Recommended)
Date Introduced2019-12-03

Regional Pricing Matrix

Live on-demand rates across all global AWS cloud regions
Lowest Price Region
$4.7210/ hr
US East (N. Virginia) (us-east-1)
Highest Price Region
$7.8000/ hr
South America (Sao Paulo) (sa-east-1)
Global Fleet Average
$5.7407/ hr
Across all active regions
Regional Coverage
22regions
Available worldwide
Linux — Available in all 22 regions
RegionOn-Demand Price
af-south-1
Africa (Cape Town)
$6.2490
ap-east-1
Asia Pacific (Hong Kong)
$7.2740
ap-south-1
Asia Pacific (Mumbai)
$4.9650
ap-northeast-2
Asia Pacific (Seoul)
$5.8120
ap-southeast-1
Asia Pacific (Singapore)
$6.3760
ap-southeast-2
Asia Pacific (Sydney)
$5.9020
ap-northeast-1
Asia Pacific (Tokyo)
$6.3760
us-gov-east-1
AWS GovCloud (US-East)
$5.9530
us-gov-west-1
AWS GovCloud (US-West)
$5.9530
ca-central-1
Canada (Central)
$5.2600
eu-central-1
EU (Frankfurt)
$5.9020
eu-west-1
EU (Ireland)
$5.2600
eu-west-2
EU (London)
$5.5300
eu-south-1
EU (Milan)
$5.5300
eu-west-3
EU (Paris)
$5.5170
eu-north-1
EU (Stockholm)
$5.0160
me-south-1
Middle East (Bahrain)
$5.7860
sa-east-1
South America (Sao Paulo)
Highest
$7.8000
us-east-1
US East (N. Virginia)
Lowest
$4.7210
us-east-2
US East (Ohio)
Lowest
$4.7210
us-west-1
US West (N. California)
$5.6710
us-west-2
US West (Oregon)
Lowest
$4.7210