Artificial Intelligence

Icon - Markets - AI - Magic

Enabling new applications

AI unleashes new waves of innovation, powering new applications and business models in multiple markets.

Icon - Markets - AI - Technologies

The right platforms

ARM ecosystem technologies for hardware and software enable the development of intelligent, distributed, heterogeneous, and secure solutions.

Icon - Markets - AI - Pervasive

Distributed intelligence

AI solutions are already used by billions of people worldwide on platforms such as smartphones, but are quickly spreading into more diverse applications.

Icon - Markets - AI - Trusted

Security in intelligence

The information, communications and inferences of an AI system must be secure for the safety and privacy of its users.

The foundation for intelligent computing

Artificial Intelligence building blocks

ARM technologies enable the world’s most popular AI platform — the smartphone — including machine learning features like predictive text, speech recognition, and computational photography. We also enable newer AI platforms like voice assistants and consumer robots that are revolutionizing how people interact with technology and the way technology interacts with the world.

As we move from the era of web services and apps to an era of AI-based systems, this revolution will happen faster. AI will be pervasive in all computing applications. Autonomous driving systems will improve the safety and change the usage patterns of our cars. The smartphone will become more intuitive and proactive. And in addition to improving existing applications, radical new ideas and businesses models will change our lives in ways we can’t yet imagine.

ARM will enable this new era through high-performing, efficient, and secure technologies.

Right for each application

Responsive and efficient AI 

ARM’s wide portfolio of CPU, GPU, and interconnect technologies power many current AI applications and also deliver on future requirements. ARM DynamIQ technology enables faster response times between the CPU and specialized accelerator hardware on a SoC (system on a chip). New processor instructions and compute libraries dedicated for machine learning (ML) and AI applications will deliver up to a 50x boost in AI performance over the next 3-5 years relative to Cortex-A73-based systems today.

Software enablement is key to AI application development. The ARM Compute Library provides optimized low-level functions to enable faster deployment of computer vision (CV), ML, and AI applications running on CPU and GPU cores. ARM is also working with its ecosystem partners to optimize AI software solutions that will enable developers to create efficient and responsive applications.

Cameras will soon replace a variety of sensors by using computer vision accelerators that can process raw video data to detect objects and some information about their state. ARM’s Spirit vision accelerator IP enables applications such as people detection and tracking, recognizing individuals and their movements throughout a space.

Distributed intelligence from the cloud to edge

Enable AI everywhere

AI solutions based on deep learning typically involve training in the cloud, with the trained networks then used locally on devices, including smartphones, smart cameras and vehicles. Distributed intelligence will eventually be in every part of the system, from the datacenter to network infrastructure, to both smart and constrained devices. Latency and network bandwidth requirements have been a catalyst for deploying intelligence at the edge. Data privacy and security concerns will further drive the training stage the same way.

ARM's common architecture will support diverse AI applications and their specific requirements everywhere they need to run. Powerful application-class cores and interconnects can enable the high compute power required in the datacenter. Highly power efficient microcontrollers can enable AI algorithms for even highly constrained, battery powered edge devices, such as wearables & sensors.

ARM enables trust in AI

The foundation of safe and secure systems

Security and privacy is of the utmost importance and ARM TrustZone technology is proven to be a key piece of technology. Edge devices will carry out more analysis, sending only essential data back to the datacenter, reducing the potential for data breach and remote monitoring. The scalability of the ARM architecture allows it to be applied in network and datacenter systems as well as the smaller edge nodes, keeping the system secure.

AI systems will be applied in potentially hazardous situations. Applications such as autonomous vehicles and robots require the highest levels of reliability and the ability to fail safely. ARM is at the forefront of developing processor and on-chip interconnect technology that supports the Functional Safety ISO 26262 standard.