Arm unveils Armv9 architecture: AI, SVE2, and improved security
Arm has announced the Armv9 architecture for next-generation devices. According to the company, Armv9 will meet "the global demand for specialized processing with enhanced security and artificial intelligence." The new Armv9 architecture will be the foundation of the next 300 billion Arm-based chips.
According to the company's website, the number of Arm-based chips continues to grow, with more than 100 billion devices shipped over the past five years. Arm believes that at the current rate, 100% of the world's data will soon be processed on Arm - either at the endpoint, on data networks, or in the cloud. Armv9 should accelerate the transition from universal computing to more specialized computing across all applications as AI, IoT, and 5G gain traction globally.
“As we look to the future that will shape AI, we must lay the foundation for advanced computing to address the unique challenges ahead,” said Simon Segars, CEO of Arm. - Armv9 is the answer. It will be at the heart of the next 300 billion Arm-based chips. ”
To power AI algorithms, Arm has partnered with Fujitsu to create Scalable Vector Extension (SVE) technology, which is used in Fugaku, the world's fastest supercomputer. Armv9 includes an improved version of the technology - SVE2; it will improve performance for machine learning and digital signal processing (DSP), mixed and virtual reality applications.
Data security will be ensured by the Arm Confidential Computing Architecture (CCA), which protects parts of code and data from access or modification during use, even from privileged software, by performing computations in a secure hardware environment. CCA allows computations to be performed on secure dynamically created regions - Realms - that any application can use. For example, in business applications, Realms can protect commercially sensitive data and code from the rest of the system while it is in use or at rest.
This focus on device protection should come as no surprise: according to company statistics, the damage incurred by companies as a result of cybercrime in 2021 will amount to about $ 6 trillion, with about 5.4 thousand attacks per month targeting IoT devices. And with 75.4 billion IoT devices expected to be online by 2025, more robust security measures are needed for future hardware.
Finally, ARMv9 will increase the performance of the next two generations of mobile SoCs by up to 30%. This will be achieved, among other things, thanks to ARM Total Compute technology, which will increase overall computing performance by purposefully optimizing hardware and software at the system level. In addition, Arm is developing a number of technologies to increase the frequency, bandwidth, and cache size, as well as reduce memory latency, to further improve the performance of Armv9-based processors.