Amazon EC2 G4 Instances
The industry’s most cost-effective GPU instance for machine learning inference and graphics-intensive applications
Amazon EC2 G4 instances deliver the industry’s most cost-effective GPU instance for deploying machine learning models in production and graphics-intensive applications. G4 instances provide the latest generation NVIDIA T4 GPUs, AWS custom Intel Cascade Lake CPUs, up to 100 Gbps of networking throughput, and up to 1.8 TB of local NVMe storage.
G4 instances are optimized for machine learning application deployments (inference), such as image classification, object detection, recommendation engines, automated speech recognition, and language translation that need access to low level GPU software libraries. These instances are also cost-effective solutions for graphics-intensive applications, such as remote graphics workstations, video transcoding and game streaming in the cloud. G4 instances are offered in different instance sizes with access to one GPU or multiple GPUs with different amounts of vCPU and memory - giving you the flexibility to pick the right instance size for your applications.
Increase performance and reduce machine learning inference costs
G4 instances are equipped with NVIDIA T4 GPUs which deliver up to 40X better low-latency throughput than CPUs, so more requests can be served in real time. Also, G4 instances are optimized to be cost-effective for machine learning inference, which can represent up to 90% of overall operational costs for machine learning initiatives.
High performance graphics
Graphics applications using G4 instances have been shown to have an up to 1.8X increase in graphics performance and up to 2X video transcoding capability over the previous generation Amazon G3 instances. The NVIDIA T4 GPU can also be used for graphics applications and 3D rendering with support for the latest APIs: DirectX 12, OpenGL 4.6, OpenCL 2.2, CUDA 10.1, and Microsoft DXR.
Cost-effective small scale training
G4 instances are also useful for small-scale/entry-level machine learning training jobs for those businesses or institutions that are less sensitive to time-to-train. G4 instances deliver up to 260 TFLOPs of FP16 performance and are a compelling solution for small-scale training jobs.
A choice of configurations
G4 instances come in several configurations offering up to 64 vCPUs, up to 4 NVIDIA T4 GPUs, and up to 256 GB of host memory so that you can right size your instance. A bare metal instance will also soon be available that offers 96 vCPUs, 8 NVIDIA T4 GPUs, and 384 GB of host memory.
Local Instance Storage
G4 instances offer up to 900 GB of local NVMe storage for fast data access, enabling customers to efficiently create photo-realistic and high-resolution 3D content for movies, games, and AR/VR experiences. The G4 bare metal instance that is coming soon will offer 1.8 TB of local NVMe storage.
Up to 100 Gbps Networking
G4 instances offer up to 50 Gbps of networking throughput to remove data transfer bottlenecks. The G4 bare metal instance that is coming soon will offer 100 Gbps of networking throughput.
Machine Learning Inference
The cost of ML inference can represent up to 90% of overall operational costs. G4 instances are an ideal solution for businesses or institutions looking for a more cost-effective platform for ML inference as well as a solution for machine learning inference applications that need direct access to GPU libraries such as, CUDA, CuDNN, and TensorRT.
Remote Graphics Workstations
Customers looking to use remote workstations in the cloud for running graphics applications, such as Autodesk Maya or 3D Studio Max, can use G4 instances to give them the flexibility to provision resources on a per-project basis and not be limited by availability of on-premises hardware.
Media and Entertainment
G4 instances can be used for post-production and video playout/broadcast as well as video encoding and transcoding. In addition, G4 instances can support AR/VR applications.
|Instance Size||vCPUs||Memory (GB)||GPU||Storage (GB)||Network Bandwidth (Gbps)||EBS Bandwidth (GBps)||On-Demand Price/hr*||1-yr Reserved Instance Effective Hourly* (Linux)||3-yr Reserved Instance Effective Hourly* (Linux)|
|Single GPU VMs||g4dn.xlarge||4||16||1||125||Up to 25||Up to 3.5||$0.526||$0.316||$0.210|
|g4dn.2xlarge||8||32||1||225||Up to 25||Up to 3.5||$0.752||$0.452||$0.300|
|g4dn.4xlarge||16||64||1||225||Up to 25||Up to 3.5||$1.204||$0.722||$0.482|
|Multi GPU VMs||g4dn.12xlarge||48||192||4||1x900||50||7||$3.912||$2.348||$1.564|
|g4dn.metal**||96||384||8||2x900||100||14||Coming soon||Coming soon||Coming soon|
* Prices shown are for US East (Northern Virginia) AWS Region. Prices for 1-year and 3-year reserved instances are for "Partial Upfront" payment options or "No Upfront" for instances without the Partial Upfront option.
** Coming soon
Amazon EC2 G4 instances are available in the US East (N. Virginia and Ohio), US West (Oregon and N. California), Canada (Central), Europe (Frankfurt, Ireland, London, Paris, and Stockholm), Asia Pacific (Mumbai, Seoul, Singapore, Sydney, and Tokyo), South America (São Paulo), and GovCloud (US-West) AWS Regions as On-Demand, Reserved, or Spot Instances.
Getting Started with G4 Instances
Deep Learning AMI and Deep Learning Containers
Amazon Deep Learning AMIs provide machine learning practitioners and researchers with the infrastructure and tools to accelerate deep learning in the cloud, at any scale. You can quickly launch Amazon EC2 instances pre-installed with popular deep learning frameworks and interfaces such as TensorFlow, PyTorch, Apache MXNet, Chainer, Gluon, Horovod, and Keras to train sophisticated, custom AI models, experiment with new algorithms, or to learn new skills and techniques. To learn more, visit the Amazon Deep Learning AMIs product page.
AWS Deep Learning Containers (AWS DL Containers) are Docker images pre-installed with deep learning frameworks to make it easy to deploy custom machine learning (ML) environments quickly by letting you skip the complicated process of building and optimizing your environments from scratch. AWS DL Containers support TensorFlow and Apache MXNet, with PyTorch and other deep learning frameworks coming soon. To learn more, visit the Amazon Deep Learning Containers product page.
NVIDIA GRID AMI
NVIDIA Marketplace offerings featuring NVIDIA Quadro Virtual Workstation software. These AMIs support running up to four 4k displays on a per GPU basis on a G4 instance. To find these AMIs, use this link: NVIDIA Marketplace offerings.
NVIDIA Marketplace offerings featuring NVIDIA vGaming drivers. These AMIs support running a single 4k display on a per GPU basis on a G4 instance. To find these AMIs, use this link: NVIDIA Marketplace offerings.
NVIDIA AMI with vGaming Driver
This download is available to AWS customers only. By downloading, you agree to use the downloaded software only to develop AMIs for use with the NVIDIA Tesla T4 hardware. Upon installation of the software, you are bound by the terms of the NVIDIA GRID Cloud End User License Agreement.
If you own GRID licenses, you should be able to use those licenses on your G4 instances. For more information, see NVIDIA GRID Software Quick Start Guide.