Business Software for AWS Neuron

AWS offers a wide range of services, including database storage, compute power, content delivery, and other functionality. This allows you to build complex applications with greater flexibility, scalability, and reliability. Amazon Web Services (AWS), the world's largest and most widely used cloud platform, offers over 175 fully featured services from more than 150 data centers worldwide. AWS is used by millions of customers, including the fastest-growing startups, large enterprises, and top government agencies, to reduce costs, be more agile, and innovate faster. AWS offers more services and features than any other cloud provider, including infrastructure technologies such as storage and databases, and emerging technologies such as machine learning, artificial intelligence, data lakes, analytics, and the Internet of Things. It is now easier, cheaper, and faster to move your existing apps to the cloud.

Amazon Elastic Container Service (Amazon ECS), is a fully managed container orchestration and management service. ECS is used by customers such as Duolingo and Samsung, GE and Cook Pad to run their most sensitive and critical mission-critical applications. It offers security, reliability and scalability. ECS is a great way to run containers for a variety of reasons. AWS Fargate is serverless compute for containers. You can also run ECS clusters with Fargate. Fargate eliminates the need for provisioning and managing servers. It allows you to specify and pay per application for resources and improves security by application isolation by design. ECS is also used extensively in Amazon to power services like Amazon SageMaker and AWS Batch. It is also used by Amazon.com's recommendation engines. ECS is extensively tested for reliability, security, and availability.

Amazon SageMaker, a fully managed service, provides data scientists and developers with the ability to quickly build, train, deploy, and deploy machine-learning (ML) models. SageMaker takes the hard work out of each step in the machine learning process, making it easier to create high-quality models. Traditional ML development can be complex, costly, and iterative. This is made worse by the lack of integrated tools to support the entire machine learning workflow. It is tedious and error-prone to combine tools and workflows. SageMaker solves the problem by combining all components needed for machine learning into a single toolset. This allows models to be produced faster and with less effort. Amazon SageMaker Studio is a web-based visual interface that allows you to perform all ML development tasks. SageMaker Studio allows you to have complete control over each step and gives you visibility.

Amazon Elastic Kubernetes Service is a fully managed Kubernetes services. EKS is trusted by customers such as Intel, Snap and Intuit. It also supports GoDaddy and Autodesk's mission-critical applications. EKS is reliable, secure, and scaleable. EKS is the best place for Kubernetes because of several reasons. AWS Fargate is serverless compute for containers that you can use to run your EKS clusters. Fargate eliminates the need for provisioning and managing servers. It allows you to specify and pay per application for resources and improves security by application isolation by design. EKS is also integrated with AWS Identity and Access Management, AWS CloudWatch, Auto Scaling Groups and AWS Identity and Access Management, IAM, and Amazon Virtual Private Cloud (VPC), allowing you to seamlessly monitor, scale, and load balance your applications.

AWS Trainium, the second-generation machine-learning (ML) accelerator, is specifically designed by AWS for deep learning training with 100B+ parameter model. Each Amazon Elastic Comput Cloud (EC2) Trn1 example deploys up to sixteen AWS Trainium accelerations to deliver a low-cost, high-performance solution for deep-learning (DL) in the cloud. The use of deep-learning is increasing, but many development teams have fixed budgets that limit the scope and frequency at which they can train to improve their models and apps. Trainium based EC2 Trn1 instance solves this challenge by delivering a faster time to train and offering up to 50% savings on cost-to-train compared to comparable Amazon EC2 instances.

Amazon Elastic Compute Cloud Trn1 instances powered by AWS Trainium are designed for high-performance deep-learning training of generative AI model, including large language models, latent diffusion models, and large language models. Trn1 instances can save you up to 50% on the cost of training compared to other Amazon EC2 instances. Trn1 instances can be used to train 100B+ parameters DL and generative AI model across a wide range of applications such as text summarizations, code generation and question answering, image generation and video generation, fraud detection, and recommendation. The AWS neuron SDK allows developers to train models on AWS trainsium (and deploy them on the AWS Inferentia chip). It integrates natively into frameworks like PyTorch and TensorFlow, so you can continue to use your existing code and workflows for training models on Trn1 instances.

Amazon EC2 Inf1 instances were designed to deliver high-performance, cost-effective machine-learning inference. Amazon EC2 Inf1 instances offer up to 2.3x higher throughput, and up to 70% less cost per inference compared with other Amazon EC2 instance. Inf1 instances are powered by up to 16 AWS inference accelerators, designed by AWS. They also feature Intel Xeon Scalable 2nd generation processors, and up to 100 Gbps of networking bandwidth, to support large-scale ML apps. These instances are perfect for deploying applications like search engines, recommendation system, computer vision and speech recognition, natural-language processing, personalization and fraud detection. Developers can deploy ML models to Inf1 instances by using the AWS Neuron SDK. This SDK integrates with popular ML Frameworks such as TensorFlow PyTorch and Apache MXNet.

Amazon EC2 instances G5 are the latest generation NVIDIA GPU instances. They can be used to run a variety of graphics-intensive applications and machine learning use cases. They offer up to 3x faster performance for graphics-intensive apps and machine learning inference, and up to 3.33x faster performance for machine learning learning training when compared to Amazon G4dn instances. Customers can use G5 instance for graphics-intensive apps such as video rendering, gaming, and remote workstations to produce high-fidelity graphics real-time. Machine learning customers can use G5 instances to get a high-performance, cost-efficient infrastructure for training and deploying larger and more sophisticated models in natural language processing, computer visualisation, and recommender engines. G5 instances offer up to three times higher graphics performance, and up to forty percent better price performance compared to G4dn instances. They have more ray tracing processor cores than any other GPU based EC2 instance.

Amazon EC2 instances P4d deliver high performance in cloud computing for machine learning applications and high-performance computing. They offer 400 Gbps networking and are powered by NVIDIA Tensor Core GPUs. P4d instances offer up to 60% less cost for training ML models. They also provide 2.5x better performance compared to the previous generation P3 and P3dn instance. P4d instances are deployed in Amazon EC2 UltraClusters which combine high-performance computing with networking and storage. Users can scale from a few NVIDIA GPUs to thousands, depending on their project requirements. Researchers, data scientists and developers can use P4d instances to build ML models to be used in a variety of applications, including natural language processing, object classification and detection, recommendation engines, and HPC applications.

Amazon EKS Anywhere, a new Amazon EKS deployment option, allows you to create and manage Kubernetes clusters anywhere. This includes on your own virtual machines (VMs), and bare metal servers. EKS Anywhere offers an easy-to-install software package that allows you to create and operate Kubernetes clusters. It also includes automation tools for supporting the cluster's lifecycle. EKS Anywhere provides a consistent AWS management experience for your data center. It builds on the strengths and capabilities of Amazon EKS Distro (the same Kubernetes which powers EKS on AWS). EKS Anywhere makes it easy to buy or build your own management tools. It allows you to create EKS Distro clusters and update software. You can also handle backup and recovery. EKS Anywhere allows you to automate cluster management, reduce support cost, and eliminate redundant effort of using multiple open-source or third-party tools to operate Kubernetes clusters. AWS fully supports EKS Anywhere.

AWS Deep Learning AMIs are a secure and curated set of frameworks, dependencies and tools that ML practitioners and researchers can use to accelerate deep learning in cloud. Amazon Machine Images (AMIs), designed for Amazon Linux and Ubuntu, come preconfigured to include TensorFlow and PyTorch. To develop advanced ML models at scale, you can validate models with millions supported virtual tests. You can speed up the installation and configuration process of AWS instances and accelerate experimentation and evaluation by using up-to-date frameworks, libraries, and Hugging Face Transformers. Advanced analytics, ML and deep learning capabilities are used to identify trends and make forecasts from disparate health data.

Amazon Elastic Compute Cloud's (Amazon EC2) instances P5 powered by NVIDIA Tensor core GPUs and P5e or P5en instances powered NVIDIA Tensor core GPUs provide the best performance in Amazon EC2 when it comes to deep learning and high-performance applications. They can help you accelerate the time to solution up to four times compared to older GPU-based EC2 instance generation, and reduce costs to train ML models up to forty percent. These instances allow you to iterate faster on your solutions and get them to market quicker. You can use P5,P5e,and P5en instances to train and deploy increasingly complex large language and diffusion models that power the most demanding generative artificial intelligent applications. These applications include speech recognition, video and image creation, code generation and question answering. These instances can be used to deploy HPC applications for pharmaceutical discovery.

Amazon EC2 capacity blocks for ML allow you to reserve accelerated compute instance in Amazon EC2 UltraClusters that are dedicated to machine learning workloads. This service supports Amazon EC2 P5en instances powered by NVIDIA Tensor Core GPUs H200, H100 and A100, as well Trn2 and TRn1 instances powered AWS Trainium. You can reserve these instances up to six months ahead of time in cluster sizes from one to sixty instances (512 GPUs, or 1,024 Trainium chip), providing flexibility for ML workloads. Reservations can be placed up to 8 weeks in advance. Capacity Blocks can be co-located in Amazon EC2 UltraClusters to provide low-latency and high-throughput connectivity for efficient distributed training. This setup provides predictable access to high performance computing resources. It allows you to plan ML application development confidently, run tests, build prototypes and accommodate future surges of demand for ML applications.

Amazon EC2 UltraClusters allow you to scale up to thousands of GPUs and machine learning accelerators such as AWS trainium, providing access to supercomputing performance on demand. They enable supercomputing to be accessible for ML, generative AI and high-performance computing through a simple, pay-as you-go model, without any setup or maintenance fees. UltraClusters are made up of thousands of accelerated EC2 instance co-located within a specific AWS Availability Zone and interconnected with Elastic Fabric Adapter networking to create a petabit scale non-blocking network. This architecture provides high-performance networking, and access to Amazon FSx, a fully-managed shared storage built on a parallel high-performance file system. It allows rapid processing of large datasets at sub-millisecond latency. EC2 UltraClusters offer scale-out capabilities to reduce training times for distributed ML workloads and tightly coupled HPC workloads.

Amazon EC2 Trn2 instances powered by AWS Trainium2 are designed for high-performance deep-learning training of generative AI model, including large language models, diffusion models, and diffusion models. They can save up to 50% on the cost of training compared to comparable Amazon EC2 Instances. Trn2 instances can support up to 16 Trainium2 accelerations, delivering up to 3 petaflops FP16/BF16 computing power and 512GB of high bandwidth memory. Trn2 instances support up to 1600 Gbps second-generation Elastic Fabric Adapter network bandwidth. NeuronLink is a high-speed nonblocking interconnect that facilitates efficient data and models parallelism. They are deployed as EC2 UltraClusters and can scale up to 30,000 Trainium2 processors interconnected by a nonblocking, petabit-scale, network, delivering six exaflops in compute performance. The AWS neuron SDK integrates with popular machine-learning frameworks such as PyTorch or TensorFlow.

Deep Learning Containers are Docker images pre-installed with the most popular deep learning frameworks. Deep Learning Containers allow you to quickly deploy custom ML environments without the need to build and optimize them from scratch. You can quickly deploy deep learning environments using prepackaged, fully tested Docker images. Integrate Amazon SageMaker, Amazon EKS and Amazon ECS to create custom ML workflows that can be used for validation, training, and deployment.