What We Do

Arm creates efficient, powerful solutions for addressing very large problems.

What We Do

IPs Not ICs

Arm designs the technology building blocks that semiconductor designers, equipment manufacturers, and others use to create silicon chips and specialized compute systems that are today powering a host of devices, applications, and services across global markets.

We offer a range of flexible licensing options, from those allowing unlimited experimentation with Arm's platform to specific product-by-product, single design uses.

Why Arm?

With billions of transistors and specialized subsystems operating at extreme speeds, processors remain some of the most complex devices ever produced. And every year the challenge becomes more difficult.

Arm technology is built on decades of research, investment, and engineering in areas such as processor verification, physical layout design, security, and debugging. Using Arm IP, helps partners get to market quickly and reliably, delivering unique system-on-chip (SoC) products that meet any performance need.

We are also dedicated to cultivating an ecosystem where competition, cooperation, and innovation all thrive. We do our best to help businesses make their Arm-based product succeed in their chosen market with targeted services like support and training. The health of the ecosystem is fundamental to our strategy: we succeed only if our customers and partners succeed.

We’re entering an era where computing must be pervasive, powerful, secure, and sustainable. Arm is designing the technology building blocks to make that possible."

Rene Haas, CEO, Arm

Case Study: Cortex-A53

Unveiled in 2012, the Arm Cortex-A53 CPU is one of our most popular designs. Semiconductor companies such as Qualcomm, Samsung, MediaTek, and NXP have incorporated it into a broad spectrum of entry-level, mid-range, and high-end SoCs.

These SoCs in turn have served as a foundation for popular and groundbreaking devices, including the most popular smart speakers, digital telescopes, smart glasses, computer vision systems for cars, smart security cameras, a space-saving smart oven, gaming controllers, a device for detecting bacteria in water, e-readers, a device that lets cars find their own parking spaces, personal firewalls, an internet access hub for isolated communities, tablets, and billions of smartphones and IoT devices.

Solving Computing's Biggest Problems

Performance, Power, and Cost

Performance-per-watt is the unsung hero of computing. Since the 1940s, the number of operations computers perform per watt of energy has approximately doubled every two years. The advances in chip design, material science, and manufacturing that have made performance gains possible have also steadily slashed the cost of computing.

Unfortunately, traditional techniques for advancing performance, power, and cost are becoming more complex and expensive. A slackening of the pace would reverberate across the technology landscape. Cellular networks would become more costly to manage. Datacenters would require more racks, equipment, and real estate. New applications would bog down devices. The beneficial cycle of smaller, cheaper, faster, and better would begin to slow down.

An Energy Crisis? Datacenter power has remained relatively flat for a decade. Some predict it could grow by up to 7x in the coming decade with AI, IoT, and other applications.

The advances in chip design, material science, and manufacturing that have made performance gains possible have also steadily slashed the cost of computing.

AI Everywhere

By 2025, around 90% of new enterprise apps will include some form of AI. We will also be surrounded by AI- and ML-enabled devices: analysts estimate there will be over 8 billion voice-assisted devices by then, or enough for one for every person on the planet.⁵

Delivering the potential of AI compute everywhere requires us to make rapid progress across areas such as processing power, system design, and neural network model development to meet advancing market needs. Arm and its ecosystem are focused on matching what the world requires, while simultaneously seeking to protect data confidentiality, privacy, and security.

The Next Frontier - Specialized Processing: Combining the Arm Cortex-M55 CPU and Arm Ethos-U55 neural processing unit (NPU) delivers a performance uplift of up to 50x for ML and 25x for efficiency.

By 2025, around 90% of new enterprise apps will include some form of AI."

Security

Maintaining robust security has become the connected computing world's greatest challenge. Hackers exposed an estimated 37 billion private records in 2020. Ransomware attacks grew by 150% with the average cost of an attack rising to $1.85 million. In all, experts expect cybercrime to cause more than $10.5 trillion worth of damages by 2025.

Consumers and businesses are struggling to keep up. Only 47% of companies carry out a full threat analysis on new products, according to a survey by PSA Certified, while 54% said an uncertain ROI and lack of buy in prevented them from investing more in continuous security measures. Markets such as IoT and AI will slow down if security problems persist.

To achieve technology's potential, we need to make security better and make it far easier to use.

To achieve technology's potential, we need to make security better and make it far easier to use.

Lacking Confidence: only 4% of organizations are completely confident in their security policy when it comes to protecting third-party data.

Armv9

Armv9, our latest architecture, will be at the forefront of the next 300 billion Arm-based chips. Some of the enhancements include:

Arm Confidential Computing Architecture (CCA)
Arm CCA is a framework for protecting code and data, even while in use, at the chip level. Realms, a feature introduced by CCA, will allow users to create isolated enclaves for exchanging information or performing transactions with trusted parties.

Memory Tagging Extension (MTE)
MTE detects latent flaws in code to eliminate future problems. More than 70% of vulnerabilities in major software platforms come from the memory issues MTE finds. MTE also improves application stability and reduces crashes.

Maximizing Performance through System Design
The Armv9 architecture, combined with Arm's Total Compute approach to system-level design (see 'Our Products'), provides a foundation for achieving annual double-digit gains in performance.

Better AI Performance and Simplified Development
Scalable Vector Extension 2 (SVE2) accelerates AI adoption by making it easier for device designers and software developers to add vector processing into their products and thereby improve AI and 'real world' signal performance.

Watch Video - The Next 300 Billion: Arm Chief Architect Richard Grisenthwaite and Arm Fellows provide a deep dive into Armv9, the foundation for the next decade of Arm devices. Armv9 chips began to sample in 2021.