Arm has announced its first 64-bit real-time processor, the Cortex-R82

Arm announced the release of the Arm Cortex-R82, Arm's first 64-bit, Linux-enabled Cortex-R processor, a real-time processor that processes data close to where it is stored and is designed to accelerate the development and deployment of next-generation enterprise and computing storage solutions.

The amount of iot data is expected to exceed 79 ZB by 2025 , and the real value of data comes from the insights it generates. In order to improve security, latency, and energy efficiency, the closer the data insights are processed to where the data was generated, the better. Computational storage has emerged as the key to data storage because it puts computing power directly on storage devices, allowing businesses to access important information securely, quickly and easily.

Neil Werdmuller, Director of storage solutions at Arm, said: "Bringing data processing closer to the data storage location requires higher performance. Depending on the workload, the new Arm Cortex-R82 delivers up to a 2x performance improvement over the previous generation Cortex-R. With optional Arm Neon technology, storage applications can run new workloads, including machine learning, with lower latency and provide additional acceleration. The Cortex-R82 itself is a 64-bit architecture with access to up to 1TB of DRAM for advanced data processing by storage applications."

Storage controllers traditionally operate through bare metal /RTOS workloads as data storage and access. The Cortex-R82 offers an optional Memory Management Unit (MMU) that enables feature-rich operating systems to run directly on the storage controller and creates opportunities for new and improved applications that benefit consumers and businesses alike.

Create value for a variety of applications

Processing data directly where it is stored opens up huge opportunities for applications such as the Internet of Things, machine learning and edge computing. This is critical in predictable storage usage scenarios, such as "database acceleration", which means that no large files are transferred and security and privacy are improved; In the application of "video transcoding", data can be transcoded or encoded efficiently for streaming, and if necessary, at different bit rates and resolutions. At the same time, it is increasingly important for applications such as transportation, for example, modern aircraft can generate several terabytes of flight data per day, which is often analyzed after unloading. Computational storage provides airlines with the ability to analyze this data in real time on hard drives, so when a plane lands, they can secure the next flight in 30 minutes or less, providing passengers with faster reprovisioning and increased security.

Flexibility is the key

As the storage market evolves, one of the biggest needs for partners is flexibility. Based on the new capabilities of the Cortex-R82 processor, partners can achieve multi-core designs of up to eight cores and adjust the type of workloads running on the storage controller based on external software requirements. For example, parking lots regularly use video surveillance to identify license plate information for subsequent charging. Car parks collect license plate data during the day, meaning most cores are used for intensive storage; At night, these cores are used to process data as a basis for charging, and can be adjusted to perform data analysis and machine learning needs. As storage controllers become more versatile to meet different market needs, the Cortex-R82 provides an extremely flexible architecture while reducing costs and time to market.

Get developers up to speed

To develop the system-level chips needed for these future use cases, Arm's partners need access to simple and cost-effective technology, and more importantly, they must confirm that the software can work. Arm is researching ways for partners to reduce complexity and cost, which is why Cortex-R82 needs to leverage the Arm Linux and server ecosystem. The ability to run Linux gives developers a whole new set of software tools and technologies for storage applications, such as Docker and Kubernetes, and provides a way to achieve acceleration. The Cortex-R82 also leverage Arm's foundation in security and is compatible with Arm TrustZone to ensure storage controller firmware can be separated from Linux or other real-time workloads.

Werdmuller added: "Arm's expertise and experience in data storage puts it in an even better position to respond to the changing needs of the market. 85% of today's hard disk controllers and solid-state drive controllers are based on the Arm architecture, Arm is already a trusted partner for billions of storage devices. In a world of billions of connected devices, data processing can't just happen in the cloud. The Cortex-R82 will help ensure that enterprises tap into insights and get the most value from their future iot deployments more efficiently and securely."