Recently, at the 9th Annual China Electronic ICT Media Forum and 2021 Industry and Technology Outlook Symposium hosted by Yiweixun, Zou Wei, senior FAE manager of Arm Technology, gave an in-depth interpretation of Arm’s newly released target after ten years of polishing for the industry. The new-generation Armv9 architecture with different levels of computing power requirements and machine learning (ML) development paths is not only another milestone in the evolution of the Arm architecture, but will also become the foundation for Arm to promote industry innovation in the next decade and beyond.
Zou Wei, Senior FAE Manager of Amou Technology, gave a live speech
Armv9: How to change the way high performance computing?
Looking at the development of semiconductor design in the past ten years, especially in the field of terminal consumption, the productivity improvement brought by computing power has put a large amount of computing power into the hands of every consumer through the mobile Internet. At the same time, processes that traditionally relied on manual data collection and analysis are evolving into machine-to-machine data collection, storage, analysis, and service in the fifth wave of computing. The resulting massive data requires more powerful computing power to be configured on the cloud and edge sides, and the obtained data is analyzed, constructed and made decisions through the data center.
In addition to becoming the mainstream architecture of the mobile wave, Arm has entered new scenarios such as edge and cloud. Its unique positioning can achieve a solid security foundation for the industry, while ensuring that developers can use Arm’s security in the easiest and fastest way technology. According to Zou Wei, by the end of 2020, a total of 180 billion chips based on Arm IP have been sold, covering every corner and every place in all walks of life. The shared data must be calculated by the Arm processor in one or more links, whether it is an IoT device for data collection, a mobile phone, a smart computer, or even a server or supercomputer for daily use.
According to Arm’s forecast, the shipments of partners based on Arm chips will exceed 300 billion in the next ten years, and 100% of the shared data will be processed on Arm chips, whether in the terminal or in the cloud. Therefore, a new processor architecture is needed to change the way computing over the next decade to deliver higher performance, improve security features, and enhance workload capabilities.
Zou Wei introduced the new Armv9 architecture features on the spot
“For this reason, Arm officially launched Armv9, a new generation of architecture for the next ten years. On the basis of Armv8, which has been widely used, it continues to use AArch64 as the benchmark instruction set to maintain backward compatibility. Improvements in security, AI and ML, as well as scalable vector scaling and DSP, expand the scope of applications.” Zou Wei said, “The processors implemented by the Armv9 architecture can be used in mobile computing, HPC high-performance computing, automotive and AI, etc. market, etc., to meet the global demand for increasingly powerful security, artificial intelligence and specialized processing, which means that computing technology based on Arm architecture will also gain leadership in markets other than smartphones, with the help of the mobile ecosystem. Huge scale advantage, creating leading solutions in application fields such as laptops, desktops, and cloud.”
Arm comprehensive computing solutions enable the ultimate digital experience
Although the global semiconductor industry has changed in recent years, providing advanced computing power has always been the fundamental goal of industrial development. From this perspective, what challenges will chip design companies face? Zou Wei analyzed that, first of all, the load of the CPU is becoming more and more complex, and the requirements will be different in different scenarios, and the challenges faced by each link are different; and although Moore’s Law has slowed down, the market’s demand for performance and power consumption is still different. The pursuit has not stopped, which is also a big challenge; in addition, the high mask price and the longer production cycle, the cost of designing a 5nm chip will be very high and the production cycle will be very long. “We want the chip design to cover not only current applications but also future applications for a longer life cycle, which requires us to address the challenges encountered by these customers through architecture, design, and IP,” he emphasized.
If empowering the ultimate digital experience is the driving force behind Arm’s vision of building the future of computing, then Arm’s latest comprehensive computing solution is the first step in the next decade of the Armv9 era. Arm comprehensive computing solutions take a system-wide holistic approach to optimization across hardware IP, physical IP, software, tools and standards to provide Arm partners with a broader choice of use cases and costs across all end-market segments range, allowing them to quickly bring high-performance products to market through the latest technology with confidence.
As the cornerstone of Arm’s comprehensive computing solutions, the new Arm Cortex CPUs include Arm’s most powerful Arm Cortex-X2, the first large-core CPU based on Armv9 architecture, Arm Cortex-A710, and Arm’s first high-end CPU in the past four years. Efficient small-core Arm Cortex-A510 and Armv9-A CPU clusters. To support the ecosystem’s need for performance, Arm’s global partners are working to ensure that all apps will support 64-bit by the end of this year. As Zou Wei said in his speech, in the era of computing power diversification, Arm believes that the computing power level of the CPU is still very important, and it has built the foundation for the digital future.
Behind the 20-30% improvement in the performance of each generation of CPU is Arm’s unremitting pursuit of a little bit of improvement in frequency, bandwidth, latency, cache, process level and other indicators, and has been leading the innovation of computing infrastructure. with the progress of the industry. Therefore, in addition to improving CPU performance, Arm’s comprehensive computing solution also hopes to make system-level improvements to bring customers a better design experience.
In addition, the visual experience is the key for consumers to interact with and enjoy the device. It is reported that the latest Arm Mali series GPUs are paired with Armv9 CPUs in a comprehensive computing solution, which can provide an excellent user experience. According to Zou Wei, the Mali series of GPUs will continue to exceed everyone’s expectations in the future, bringing ray tracing technology, variable rate shading VSR technology and other advanced rendering technologies.
Arm’s interconnect technology is also critical to improving system performance. The latest CoreLink CI-700 coherent interconnect technology and CoreLink NI-700 network-on-chip interconnect technology work seamlessly with Arm CPU, GPU and NPU IP to enhance system performance across SoC solutions. The CoreLink CI-700 and CoreLink NI-700 provide hardware-level support for new Armv9-A features, along with support for higher security, improved bandwidth and latency.
Introduce new security features to protect data core assets
With the proliferation of connected devices around the world, the security of chips is increasingly important. The fundamental need for security is to rethink how security is handled, and the Armv9 architecture redesigns the way security applications work by introducing the Arm Confidential Compute Architecture (CCA). Confidential computing performs computations by creating a secure hardware-based operating environment that protects portions of code and data from being accessed or modified, even by privileged software.
Zou Wei further explained that Arm CCA will introduce the concept of dynamically creating secret realms (Realms). The secret realm is oriented to all applications and runs in an environment independent of the secure or non-secure environment. It is a dynamic security for storing data and executing code. A zone, separate from the privileged mode of the operating system or hypervisor, for the purpose of protecting data security. For example, in commercial applications, a confidential domain can protect commercially confidential data and code in a system, whether they are in use, idle, or in transit. Built on Arm’s TrustZone technology, the Arm Confidential Computing Architecture is a hardware version of software containers that allow applications to easily run on different systems. For example, applications running in non-secure areas can dynamically apply for confidential areas to protect their algorithms or data.
“Arm also worked with partners to develop the platform security architecture PSA, as a set of standard threat models, measures and timings for terminal designers to use, which, combined with hierarchical certification, allows purchasers of this equipment to ensure asset security. For example, in China, we have cooperated with the China Academy of Information and Communications Technology’s Taier Laboratory to support the large-scale deployment of secure IoT solutions based on the PSA framework.” Zou Wei added.
Building a large computing platform in the new computing era
In addition to security, machine learning is also seen as an important part of Armv9, as more and more machine learning workloads will become commonplace in the next few years, in terms of machine learning Armv9 architecture supports BFloat16 format for better to support Int8 computing and BFloat16? machine learning; the introduction of Scalable Vector Extension 2 (SVE2) can better help developers develop high-level application scenarios, dealing with 5G, virtual reality and augmented reality and Large gains for workloads such as image and speech recognition.
Looking at the ever-changing digital world, smart terminals not only need to be flexible and low-power, but also meet the workload requirements of AI/ML. These challenges are driving the development of more secure and dedicated processing, but also opening new markets, The key to seizing new opportunities. The release of the Armv9 architecture and comprehensive computing solutions based on security technologies will unlock new experiences across the ecosystem and build trusted digital services for the next decade.
Zou Wei concluded: “The next-generation Armv9 architecture will enable developers to build and program the trusted computing platform of the future by bridging the critical gap between hardware and software, helping our partners achieve faster time-to-market and cost control. Amou China is committed to comprehensively promoting the computing power revolution from technology to ecology, and will increase investment in self-research architecture, continue to work with ecological partners, and rely on the leading Arm global ecosystem to deepen and strengthen local R&D. A corporate strategy that empowers innovation in local industries.”