Announcing the preview of new Azure VMs based on the Azure Cobalt 100 processor
Published May 21 2024 08:29 AM 21.6K Views
Microsoft

Today, Microsoft is announcing the preview of the new Azure Virtual Machines (VMs) featuring the Azure Cobalt 100 Arm-based processor. The Cobalt 100 processor is based on the Neoverse N-series (N2) Arm CPU design, which is optimized for the performance of scale out cloud-based applications. The preview includes the general purpose (Dpsv6-series and Dplsv6-series) and memory optimized (Epsv6-series) VM series.

To request access to preview, please fill out this form.

 

Cobalt 100 chip image.png

 

The new Dpsv6 and Dpdsv6 general purpose VMs are engineered to efficiently run scale-out workloads and cloud native computing solutions such as those running on Azure Kubernetes Service (AKS). These VMs are ideal for small to medium open-source databases, application servers and web servers. These VMs also excel with containerized applications and can be leveraged by Arm developers in CI/CD pipelines, development and test scenarios. The new Dplsv6 and Dpldsv6 VMs are ideal for media encoding, small databases, gaming servers, microservices and any workloads that do not require higher RAM per vCPU. Additionally, the Epsv6 and Epdsv6 memory-optimized VMs have higher memory per vCPU to meet the requirements of large databases, in-memory caching applications and data analytics workloads. All the new Cobalt 100 VMs enable customers to seamlessly run modern, dynamic, scalable applications.

 

The Azure Cobalt 100 VMs can deliver up to 1.4x CPU performance, up to 1.5x performance on Java-based workloads and up to 2x performance on web servers, .NET applications and in-memory cache applications compared to previous generation Azure Arm-based VMs. These VMs support 4x local storage IOPS (with NVMe direct) and up to 1.5x network bandwidth compared to the previous generation Azure Arm-based VMs.

 

You can select from a range of Azure virtual machines that meet the CPU performance and memory needs of your workloads. You can choose from three memory ratios for a given vCPU size, giving you the flexibility to select the configuration that works best for your workload.

  • The new Dpsv6-series offers up to 96 vCPUs with 384 GiBs of RAM (4:1 memory-to-vCPU ratio).
  • The new Dplsv6-series offers up to 96 vCPUs with 192 GiBs of RAM (2:1 memory-to-vCPU ratio).
  • The new Epsv6-series offer up to 96 vCPUs with up to 672 GiBs of RAM (up to 8:1 memory-to-vCPU ratio) for more memory intensive workloads.

All these VM series are available with and without a local disk so that customers can choose the option that best fits each workload.

 

You can deploy these new VMs using existing methods including the Azure portal, SDKs, APIs, PowerShell, and the command-line interface (CLI).

 

During the preview period, the VMs are available in the Central US, East US, East US 2, North Europe, Southeast Asia, West Europe, and West US 2 Azure regions. The number of regions will continue to expand in 2024 and beyond. Explore future regional availability options by visiting the Azure product availability page.

 

During the preview, the new Azure Cobalt 100 VMs are available for free. However, your Azure subscription will be billed for fees and applicable taxes associated with use of other Azure resources and services, such as disk storage, used in the deployment of these preview VMs.

 

The new virtual machines support all remote disk types such as Standard SSD, Standard HDD, Premium SSD and Ultra Disk storage. To learn more about various disk types and their regional availability, please refer to Azure managed disk type. Disk storage is billed separately from virtual machines.

 

The new Azure Cobalt 100 VMs support a wide range of Linux OS distributions including Canonical Ubuntu, CentOS, Debian, Red Hat Enterprise Linux, SUSE Enterprise Linux, Alma Linux, Azure Linux (via AKS), Flatcar Linux and more. Client application developers can take advantage of Azure’s highly available, scalable, and secure platform to run cloud-based builds and test workflows. For Windows developers, Insider Preview of Windows 11 Pro and Enterprise are available for Azure Cobalt 100 VMs. Customers can access the full list of images in the Azure Marketplace.

 

Microsoft has decades of experience with Arm-based technologies through collaborations with software and hardware partners. The preview of the Azure Cobalt 100 VMs continues to showcase Microsoft’s commitment to developing a vibrant Arm ecosystem to accelerate customer innovation and in helping customers build great solutions. Major developer platforms and languages, such as .NET, C++ and Java, provide native Arm support to take advantage of the benefits that this processor architecture brings.

 

Microsoft continues to invest in optimizing these developer platforms and languages on Linux and Windows to leverage the capabilities of the latest Arm architecture. .NET 8 adds numerous optimizations for Arm and C++ similarly brings its share of Arm optimizations with the Visual Studio 17.8 release. Visual Studio 17.10 also introduces SQL Server Data Tools (SSDT) for the Arm native Visual Studio. For Java customers, the Microsoft Build of OpenJDK has been built with Arm compiler optimizations and has been certified on Linux and Windows on Arm architectures.

GitHub Actions, GitHub's CI/CD workflow engine, is an integral part of many developers’ workflows, and is used to continuously build, test, and deploy apps. GitHub Actions is now available for Windows and Linux on Arm in 2 flavors – self-hosted runners that can be hosted on an Arm VM or Arm device, and GitHub hosted runners that is available in private beta with GA expected later this summer.

 

Containers are a popular deployment target for many reasons - a streamlined development workflow, isolation and security, efficient resource utilization, portability, and reproducibility.  Emphasizing its commitment to developer productivity, Docker is now investing in this area by ensuring that Docker Desktop runs natively on Windows on Arm.

 

The rich ecosystem of applications and libraries, compiled natively to Windows on Arm, continues to grow. In addition to the Microsoft products and tools optimized for Arm64, developer and creator tools such as the Unity games editor, Blender, Docker, GIMP and important libraries such as Qt, are all on track to deliver Arm-native versions this year.

 

Our customers have shared their perspective:

 

"Ansys and Microsoft have a long history of collaboration, and we are excited to continue our joint work to advance semiconductor design via the Cobalt 100 chip and the power of Microsoft Azure," said John Lee, vice president and general manager of the semiconductor, electronics, and optics business unit at Ansys. "Through the combined strengths of Ansys simulation expertise and Microsoft's prowess in enabling EDA technologies, Cobalt 100-based solutions will empower designers to economically achieve optimal Power, Performance, and Area in their designs. Together, we will help propel the semiconductor industry into a new era of unparalleled efficiency and productivity.”

 

“We’ve helped thousands of customers in EDA and Systems combine the performance and scalability of the cloud providing ease-of-use and instant access to our powerful computational software, which speeds time-to-market window for innovative designs. We are excited to work with Microsoft Azure on the new Arm-based Azure Cobalt 100 to provide our mutual customers with optimal performance as they tackle the ever-increasing demands of compute, memory capacity, and price performance for gigascale, advanced-node designs.” - Mahesh Turaga, VP of Cloud Business Development, Cadence.

 

“At Elastic, we love working with the Microsoft teams, from silicon to models, said Shay Banon, co-founder and chief technology officer at Elastic. The rate of progress on the Azure team is impressive, and we are excited to collaborate with them to bring these benefits to our users as fast as possible.” 

 

“At MongoDB, we continuously explore cutting-edge technologies to enhance our solutions. We are currently evaluating Arm for its benefits and price performance, with promising results so far. We have also evaluated the Azure Cobalt 100 VM and are pleased with its capabilities, particularly in handling intensive workloads. We look forward to taking advantage of these advancements to provide best-in-class performance and even better experiences for our customers.” - Andrew Davidson, SVP, Products, MongoDB.

 

"Semiconductor companies must balance the competing priorities of schedule, cost and performance as they develop their products in sustainable and efficient ways.  An important aspect of doing this successfully is carefully matching software workloads with the appropriate computing platforms.  The combination of Microsoft Azure Cobalt 100 Arm-based VMs with Siemens EDA software will provide our mutual customers with an excellent option to meet both their product development and corporate objectives.  We continue to partner with Microsoft to create innovative solutions for the semiconductor industry.” - Craig Johnson, Vice President, Siemens EDA Cloud Solutions.

 

Snowflake, as an early adopter of Azure’s Arm technology, has successfully leveraged Arm-based VMs in Azure for our data warehouse workloads. We are now excited to elevate our performance further with Azure's Cobalt 100 VMs, anticipating significant enhancements in delivering superior customer experience.” - Gabe Bryant, Director of Engineering, Snowflake.

 

Teradata is thrilled to partner with Microsoft Azure, and their launch of the Cobalt 100 VMs.  The density and efficiency benefits will further Teradata's existing price/performance advantage for AI and data-driven workloads, scaled on Teradata’s VantageCloud Lake offering” - Daniel Spurling, SVP, Product Management, Teradata.

 

Here’s what our technology partners are saying:

 

“Customized silicon plays a fundamental role in powering the AI-accelerated workloads of the future. This suite of Azure VMs, powered by the Neoverse CSS -based Azure Cobalt 100, marks a key milestone in our longstanding partnership with Microsoft to unlock new paths to performant, efficient computing and provides Azure users more choice and greater ability to innovate using the vast Arm software ecosystem.” -Dermot O’Driscoll, vice president of product solutions, Infrastructure Line of Business, Arm.

 

The adoption of Arm-based architectures on Microsoft Azure drives cost efficiency and significant performance enhancements. With the introduction of the Azure Cobalt 100 Arm-based VMs, we are excited that Ubuntu users will see up to a 40% increase in workload performance. Our collaboration with Microsoft ensures that users can fully capitalize on these advancements with Ubuntu and Ubuntu Pro, optimizing for diverse and demanding workloads including application servers, machine learning platforms, open-source databases, in-memory caches, and cloud-native applications on AKS. This collaboration underscores our dedication to providing robust and scalable solutions that advance the capabilities of developers and enterprises alike.” - Alexander Gallagher, Vice President of Public Cloud, Canonical.

 

Whether on Azure or in their own datacenter, Red Hat is committed to providing customers with the broadest possible choice in architectures across the hybrid cloud that best meet their unique business requirements. The Cobalt 100 Arm-based Azure VMs offer a new, powerful option for customers to potentially drive greater efficiencies and scale in their operations, and we look forward to exploring how we can best enable these Arm-based processors on Red Hat Enterprise Linux.” - Ronald Pacheco, senior director, Red Hat Enterprise Linux Product and Ecosystem Strategy, Red Hat.

 

You can learn more about the new Azure Cobalt 100 VMs by visiting the specification pages: Dpsv6-series, Dpdsv6-series, Dplsv6-series, Dpldsv6-series, Epsv6-seriesEpdsv6-series.

 

Have any questions? Please reach us at Azure Support and our experts will be there to help you with your Azure journey.

 

Additional resources: