This page was machine-translated and may differ from the original. View original
Arm Announces Mali Multimedia IP Family for Universalizing Powerful Visual Experiences
You can enjoy advanced features supporting machine learning, HDR, and complex UI.
Arm has announced the Mali Multimedia Suite, which provides a powerful visual experience even on entry-level mobile devices. This enables more mobile users to enjoy advanced features that support machine learning (ML), HDR, and complex UI.
The content that smartphones must process is becoming increasingly complex. Furthermore, consumers are naturally demanding more from entry-level and low-cost devices, making richer, multi-layered user interfaces and extensive support for the latest applications and technologies essential.
Furthermore, ML is no longer a domain exclusive to premium smartphones. Arm is well aware of this trend, as over 90% of currently shipped AI-enabled devices are based on Arm architecture. All users expect to easily use applications that leverage ML, and gaming is expanding the reach of entry-level mobile devices, led by 3D games, mixed reality (MR), and 4K content, which are more widely used than ever before.
Arm announced the Mali multimedia product family, an innovative solution capable of meeting the requirements for these next-generation experiences. The new IP family, consisting of video, display, and graphics processors, enables seamless integration with existing DynamIQ-based CPUs and other Arm IP, completing Arm's next-generation solutions for entry-level mobile and DTV.
Based on the previous model, the Mali-G51, the Mali-G52 is designed to enable more complex graphics, allowing more ML functions to be implemented in entry-level mobile systems with limited power and bandwidth.
The Mali-G52 utilizes a more extensive execution engine with 8 pipelines, twice as many as the previous model which used 4 pipelines, providing more powerful graphics performance in the same semiconductor size and boasting 30% improved performance per unit area.
With 15% higher energy efficiency, it reduces device power consumption and heat generation, enabling games using battery-intensive technologies like AR to run for longer periods. With 3.6 times higher ML performance compared to the previous generation, it enables all types of devices to support next-generation ML use cases.
Computing requirements for all devices are increasing, and users expect a certain level of content quality and highly complex DTV UIs even on mid-range devices. We aim to meet these demands through the Arm Mali-G31.
The Mali-G31 is the first ultra-efficient GPU designed based on the innovative Bifrost architecture. As Arm's ultra-compact processor supporting OpenGL ES 3.2 and the latest Vulkan API, it enables developers to develop on millions of devices. It features a 20% reduction in size and a 20% improvement in performance density compared to the previous Bifrost model, Mali-G51, providing excellent energy efficiency while reducing chip area.
The Mali-D51 packs the key features of the Mali-D71, a premium display processor released in 2017, into Arm's smallest DPU. The first entry-level display processor to be released is designed based on the Komeda architecture and boasts the following performance.
The Mali-D51, optimized to work seamlessly with other IPs in the Mali multimedia family, enables HDR support even on entry-level devices when integrated with Assertive Display 5 and improves memory management efficiency across the system when used with the CoreLink MMU-600.
Ian Smythe, Senior Director of Arm’s Client Business Unit, explained, “Arm’s partners have shipped 125 billion chips to date, and we expect this figure to increase to 200 billion in 2021. This is possible because Arm is constantly striving for innovation and is committed to providing scalable CPU, GPU, video, and display architectures across a wide range of devices in the computing industry.” He added, “These principles are also clearly evident in the Mali multimedia family, which Arm announced to provide powerful visual experiences to a wider audience.”
본 기사에 대한 정정·반론·추후보도 청구는 보도 청구 안내를, 그간 게재된 보도문은 정정·반론보도 모아보기를 참고해 주세요.













