OpenGL 4.3 support improves graphics compatibility and performance, while new AI acceleration helps developers get more from Apple silicon
Accelerate Local AI Development
Run More Professional Windows Applications on Mac
AUSTIN, Texas, Aug. 25, 2026 (GLOBE NEWSWIRE) — Parallels, a global leader in cross-platform virtualization solutions, today announced Parallels Desktop 27, expanding the range of professional Windows applications users can run on Apple silicon Macs through a new Metal-based graphics driver with OpenGL 4.3 support. This release also introduces AI acceleration on Apple silicon, helping developers build and test AI workloads locally with faster machine learning performance.
Parallels Desktop is the only third-party virtualization solution currently authorized by Microsoft to run Windows in a virtualized environment on Apple silicon, making it the ideal choice for users looking to combine the flexibility of Windows with the power of Mac.
“As customers continue to push the limits of what they do on Mac, they’re looking for virtualization that keeps pace with increasingly demanding applications and workflows,” said Elena Koryakina, Chief Product and Technology Officer at Parallels. “Parallels Desktop 27 enables customers to run more of the Windows applications they depend on while giving developers faster local AI performance and expanding what’s possible on Apple silicon.”
Run More Professional Windows Applications on Mac
Parallels Desktop 27 introduces a new Metal-based graphics driver that brings hardware-accelerated OpenGL 4.3 support to Windows virtual machines on Apple silicon, a capability that Parallels believes is unique among available Mac virtualization solutions. The new driver meets the minimum OpenGL graphics level requirements for more professional Windows applications, including ArcGIS Pro, ZWCAD, and Blender 4.3, expanding compatibility while improving graphics performance for users in engineering, architecture, GIS, design, and other technical industries.
Highlights include:
- New Metal-based OpenGL 4.3 graphics driver
- Up to 160% faster OpenGL graphics performance on Macs with M3 chips and newer
- Up to 35% faster 3D rendering in ArcGIS Pro, reducing overall workflow time by up to 24%*
Accelerate Local AI Development
As developers increasingly build and test AI workloads locally, Parallels Desktop 27 introduces support for Scalable Matrix Extension (SME) instructions in Linux and Windows virtual machines running on M4 and newer Macs. The result is faster local AI development and machine learning workflows without relying on cloud infrastructure.
New developer capabilities include:
- SME2 instruction support for Linux and Windows VMs on M4 and newer Macs
- Up to 7x faster AI and machine learning matrix operations
- Up to 1.75x faster neural network inference**
- Improved performance for AI tools used in VMs such as Ollama and llama.cpp
Enterprise Deployment and Management
Parallels Desktop for Enterprise provides the deployment, security, and management capabilities IT teams need to manage virtual environments at scale. With Parallels Desktop 27, enterprise customers can use the command-line interface (CLI) to identify whether Parallels Desktop was activated using a license key or single sign-on (SSO). This enables organizations using Jamf to build extension attributes that simplify device management and reporting.
Parallels Desktop has also renewed its SOC 2 compliance, providing continued third-party assurance around the controls and processes that support the security and reliability of Parallels products and services.
Enterprise customers also benefit from Secure Golden Images Provisioning, which supports the secure distribution of approved virtual-machine images through enterprise storage, including Azure Blob Storage and SharePoint, alongside Improved Declarative Deployment for fully provisioned zero-touch Windows virtual machines. Together, these capabilities help organizations enhance their security posture, simplify deployment, and reduce the complexity of managing virtualization-enabled workflows at scale.
Built for DevOps Teams
Parallels Desktop includes capabilities for development teams building and managing CI/CD infrastructure. A new native graphical interface for the Parallels DevOps Service simplifies management of hosts, virtual machines, and automation workflows without relying entirely on command-line tools and RESTful API. Jenkins dynamic provisioning further streamlines build and test automation by enabling dynamic provisioning of Mac-hosted virtual machines for development workflows.
For more information about Parallels Desktop for Enterprise, visit www.parallels.com/products/desktop/enterprise/.
Learn more about Parallels Desktop 27 and its new features at www.parallels.com/blogs/parallels-desktop-27/.
To see if your Windows applications are compatible with Parallels Desktop, search our app database of more than 200,000 supported apps at www.parallels.com/apps
Parallels Desktop for Mac is an essential virtualization tool for professionals, developers, IT teams, and individuals, whether it is for accessing Windows applications, developing or testing software, or to run multiple operating systems (such as Windows, Linux, or other macOS versions) simultaneously on Mac computers. For more information about Parallels Desktop 27 for Mac, visit www.parallels.com.
Trial and Availability
Parallels Desktop 27 for Mac is designed exclusively for Apple silicon Macs and is available in Standard, Pro, Business, and Enterprise editions. It can be purchased or upgraded at www.parallels.com or from authorized resellers worldwide.
For more information about Parallels products, to download a free trial, or to purchase a subscription, please visit www.parallels.com.
*ArcGIS Pro testing measured up to 35% faster performance in selected tasks and up to a 24% reduction in overall workflow time with the new Metal-based OpenGL 4.3 driver.
**AI and machine-learning tests were conducted in an Ubuntu Linux virtual machine on an Apple M4 Pro Mac. Results included matrix computations up to seven times faster and neural-network inference up to 1.75 times faster. The feature requires Apple M4 hardware or newer.
Performance results are based on internal testing conducted by Parallels under controlled conditions. Performance results may vary depending on hardware, software configuration, workload, usage patterns, and other factors.
About Parallels
Parallels is a global leading brand in cross-platform solutions that make it simple for businesses and individuals to use and access the applications and files they need on any device or operating system. Parallels helps customers leverage the best technology out there, whether it’s Windows, Mac, ChromeOS, iOS, Android, or the cloud. Parallels solves complex engineering and user-experience problems by making it simple and cost-effective for businesses and individual customers to use applications anywhere, anytime. For more information, please visit www.parallels.com.
© 2026 Parallels International GmbH. All rights reserved. Parallels is a trademark or registered trademark of Parallels International GmbH in Canada, the United States and/or elsewhere. Mac and macOS are trademarks of Apple Inc. Android and ChromeOS are trademarks of Google LLC. Windows, PowerShell, SharePoint, Azure Blob Storage Microsoft Azure, and Azure Virtual Desktop are trademarks of Microsoft Corporation. ArcGIS Pro is a trademark of Esri. ZWCAD is a trademark of ZWSOFT. Blender is a trademark of the Blender Foundation. Jenkins is a trademark of the Continuous Delivery Foundation. All other company, product and service names, logos, brands and any registered or unregistered trademarks mentioned are used for identification purposes only and remain the exclusive property of their respective owners. For all notices and legal information please visit www.parallels.com/about/legal/
Ashley Ruess
[email protected]
Photos accompanying this announcement are available at
https://www.globenewswire.com/NewsRoom/AttachmentNg/6ae9b593-55b5-48f7-9b0a-6fafc315f643
https://www.globenewswire.com/NewsRoom/AttachmentNg/fa4d83df-ed51-46db-9a72-b2357a7d175d
