19:07Machine-translated from Azerbaijani by AIRead in Azerbaijani
Parallels Desktop 27 Unveiled
Parallels has unveiled Parallels Desktop 27, which allows Windows and Linux operating systems to run virtually on Mac computers. The new version's main improvements are enhanced graphics performance and accelerated artificial intelligence tasks. Parallels Desktop 27 uses a new Metal-based graphics driver for Windows virtual machines. As a result, the software now supports OpenGL 4.3, expanding the ability to run more demanding Windows applications such as Blender 4.3, ArcGIS Pro and ZWCAD on a Mac.
According to the company, OpenGL performance on Macs equipped with M3 and newer Apple processors can increase by up to 160% compared with Parallels Desktop 26. It is also reported that 3D rendering performance in ArcGIS Pro has increased by up to 35%. Another major new feature in Parallels Desktop 27 is support for Scalable Matrix Extension (SME) technology on Macs with M4 and newer processors. This capability can be used in both Windows and Linux virtual machines. SME is intended particularly for developers working locally with artificial intelligence models. Parallels says that when using tools such as Ollama and llama.cpp, the performance of matrix operations for AI and machine learning can increase by up to 7 times compared with the previous version, while neural network inference speeds can increase by up to 1.75 times. The new version also expands management capabilities for businesses. In particular, additional tools have been introduced to help organizations using Jamf manage multiple Parallels Desktop installations.
According to the company, OpenGL performance on Macs equipped with M3 and newer Apple processors can increase by up to 160% compared with Parallels Desktop 26. It is also reported that 3D rendering performance in ArcGIS Pro has increased by up to 35%. Another major new feature in Parallels Desktop 27 is support for Scalable Matrix Extension (SME) technology on Macs with M4 and newer processors. This capability can be used in both Windows and Linux virtual machines. SME is intended particularly for developers working locally with artificial intelligence models. Parallels says that when using tools such as Ollama and llama.cpp, the performance of matrix operations for AI and machine learning can increase by up to 7 times compared with the previous version, while neural network inference speeds can increase by up to 1.75 times. The new version also expands management capabilities for businesses. In particular, additional tools have been introduced to help organizations using Jamf manage multiple Parallels Desktop installations.