4 Ways Graviton Processors Make the Cloud Journey Better

May 17 2021
image

By Vyacheslav Gorlov
Senior Solutions Architect, ClearScale

Server processors power all the cloud services, and Intel and AMD have long dominated this market. Slowly, however, other companies started making inroads ─ including Amazon Web Services (AWS) with its Graviton processors. That’s why it was big news earlier this year when AWS announced it was adding to its family of Graviton2-powered Elastic Cloud Compute (EC2) instances.

It signifies that AWS is making a strong commitment to its custom-built, Arm-based Graviton server processor while continuing to broaden its portfolio of compute instances, which already features many powered by next-generation Intel and AMD silicons. It also means more choice, potential cost savings, and many other benefits for organizations using cloud services.

So, what are Graviton processors and how do they benefit organizations moving to the cloud or already using cloud services?

The Graviton Story

AWS already creates much of the hardware underlying its cloud services, so it’s not surprising the company chose to develop its own server processors. AWS custom builds its Graviton CPUs around cores that employ Arm’s Neoverse microarchitecture. Arm, considered the world’s leading semiconductor IP company, developed its Neoverse platform to deliver cutting-edge performance per watt for building new-generation cloud-to-edge infrastructure.

AWS introduced EC2 A1 instances powered by its Graviton processors in 2018, becoming the first of the cloud providers to offer Arm-based instances. The instances were developed to deliver significant cost savings over general-purpose instances for scale-out applications such as web servers, containerized microservices, and other workloads that can embrace the differences amid Arm and x86 instructions.

Then in 2020, AWS upped its game with the general availability of three EC2 instance types based on its significantly more powerful Graviton2: M6g for general-purpose workloads, C6g for compute-intensive tasks, and R6g for memory-demanding jobs.

With significantly more performance and capabilities than AWS’s first-generation Graviton processors, the Graviton2 provides up to 40% better price performance over comparable current-generation x86-based instances for a wide variety of workloads, ranging from HPC and CPU-based Machine Learning inference to open-source databases and in-memory caches.

The Benefits of Graviton Processors for Cloud Computing

Artificial intelligence (AI), data analytics, Internet of Things (IoT) devices, and other technologies are creating increasingly resource-intensive cloud workloads that require greater performance, scalability, and security. They are also driving up cloud costs, spurring a need for more cost-efficient cloud computing options. The Graviton processors deliver on both fronts. Here are four of its key benefits:

1. Best-in-Class Performance for Less

EC2 instances powered by Arm-based Graviton2 processors offer up to 50% more NVMe storage GB/vCPU over comparable x86-based. In contrast to the instances on top of the first-generation Graviton, they also deliver approximately seven times more performance, four times more compute cores, and five times faster memory.

AWS’s Graviton2-powered EC2 M6gd, C6gd, and R6gd instances, built on the AWS Nitro System, also provide up to 3.8 TB of NVMe-based SSD storage and up to 25 Gbps of network bandwidth. All AWS Graviton2-based instances support up to 19 Gbps Elastic Block Store (EBS) bandwidth. With discrete GPU support (Amazon SageMaker instances with AWS Graviton2 CPUs and NVIDIA GPUs) expected in the second half of 2021 that will have significant benefits for gaming and Machine Learning applications, efficiency figures are only going to grow.

2. Extensive Ecosystem Support

Graviton-based instances have been around for three years now, and there’s a broad Arm-centric ecosystem that includes Amazon Linux 2, Red Hat, Ubuntu, and SUSE. Arm developers can leverage this ecosystem to build applications natively in the cloud, eliminating the need for error-prone, time-consuming emulation and cross-compilation.

In addition, many of the most popular applications and services from AWS and Independent Software Vendors (ISVs), including those used in Machine Learning and other resource-intensive applications, also work with Graviton2-powered instances. Among them: Amazon ECS, Amazon EKS, Amazon ECR, Amazon CodeBuild, Amazon CodeCommit, Amazon CodePipeline, Amazon CodeDeploy, Amazon CloudWatch, Docker, Jenkins, NGINX, NodeJS, and TensorFlow.

Not surprisingly, AWS offers the same level of support for its Arm-based Graviton silicons as it does for processors built on x86 architecture. Along with a wealth of community support, it makes using Graviton-powered instances easier with fewer operational risks.

Arm itself is known and beloved over the last 35 years for its role as the main driver behind the smartphones and embeddable devices boom. When this architecture came to the cloud, both the experts and user community were already there and ready to start building the next amazing applications using tools they were already familiar with.

3. Enhanced Security

Security in the cloud continues to be a concern for many organizations, but AWS’s Graviton2 processors help reduce some of the issues. Graviton2 offers always-on 256-bit DRAM encryption and 50% faster per core encryption performance compared to silicons of the previous generation. They’re built on the AWS Nitro System, which provides enhanced security that continuously monitors, protects, and verifies the instances' hardware and firmware. That includes the Nitro security chip with dedicated hardware and software for security functions, and support for encrypted EBS storage volumes by default.

The Graviton2 processors also benefit from a range of architecture security features developed by Arm to defend against control-flow, data-access, and side-channel attacks against speculative execution.

4. Write Code Once. Use It More.

The Arm architecture of the Graviton2 processors is responsible for several benefits that reduce cloud app development time. That includes the ability to reuse the business logic built for desktop and mobile applications, reducing the effort required to support two similar yet distinct apps. All that is required to migrate server-side workloads is to recompile code for the Arm architecture that is naturally supported in modern languages and frameworks like Python/Django, PHP/Symfony, or C#/.NET.

For managed services like Amazon RDS, nothing has to be done with the code when modernizing a cluster from 4th to 5th generation instances. Just select the *g instance type, run Multi-AZ failover, and then allow your app to auto-connect using the same settings as before. With no need for routine, often error-prone activities such as syncing the logic between platforms, development teams can focus on achieving the highest levels of user satisfaction and, consequently, business value.

What’s Next: Continuous Evolution

Both AWS and Arm are recognized for innovation, so it makes sense that AWS’s Arm-based Graviton processors will continue to evolve to serve more use cases and deliver even more benefits for cloud-native application development and cloud usage in general. Innovation will also be seen among other companies developing silicons to meet the ever-increasing demands for higher performance, lower costs, and greater security for cloud apps.

What will be key for organizations that want to leverage the benefits of next-generation processors, whether they’re from AWS, Intel, or AMD, is to work with a team that stays on top of that evolution.

That describes ClearScale. We remain on the cutting edge of emerging technologies and best practices to deliver optimal cloud solutions. As an AWS Premier Consulting Partner with 11 competencies including Migrations, we have particularly strong expertise in working in AWS and with AWS’s vast portfolio of services and platforms.

Regardless of the processor powering the services required in your cloud solution, who can serve you better than ClearScale, a service provider that can draw upon industry best practices and its own deep-seated experience and expertise to deliver the solutions you need?

Ready to learn more? Schedule a free consultation with our AWS experts.

Follow ClearScale on LinkedIn, Twitter, and YouTube.

Get in touch today to speak with a cloud expert and discuss how we can help:

Call us at 1-800-591-0442
Send us an email at sales@clearscale.com
Fill out a Contact Form
Read our Customer Case Studies

Share