agsandrew -

Can Nvidia expand software business through £31bn Arm deal?

Nvidia is to buy rival chip designer Arm for £31bn in a deal it hopes will put it at the forefront of artificial intelligence for edge computing

This article can also be found in the Premium Editorial Download: Computer Weekly: Can Arm stay strong under Nvidia?

After weeks of speculation, Nvidia has agreed to purchase Cambridge-headquartered Arm for $40bn (£31bn) from Japanese conglomerate SoftBank Group.

SoftBank acquired Arm in 2016 and made a commitment to continue investing in the chipmaker until 2021, but the company recently divested its $21bn stake in T-Mobile US and has also been looking to offload Arm.

Nvidia said the acquisition of Arm would enable it to create the premier computing company for the age of artificial intelligence (AI), accelerating innovation while expanding into large, high-growth markets.

“We are acquiring one of the greatest tech companies the world has ever seen,” said Nvidia CEO Jensen Huang. “We want to grow Arm and help it become greater. We want more R&D, not less, and we want that work to be done in Cambridge.”

Arm CEO Simon Segars said: “Arm and Nvidia share a vision and passion that ubiquitous, energy-efficient computing will help address the world’s most pressing issues from climate change to healthcare, from agriculture to education. Delivering on this vision requires new approaches to hardware and software and a long-term commitment to research and development.”

In a letter to staff announcing the deal, Huang wrote: “We are joining arms with Arm to create the leading computing company for the age of AI. AI is the most powerful technology force of our time. Learning from data, AI supercomputers can write software no human can. Amazingly, AI software can perceive its environment, infer the best plan, and act intelligently. This new form of software will expand computing to every corner of the globe. Someday, trillions of computers running AI will create a new internet – the internet of things – thousands of times bigger than today’s internet of people.”

“We are acquiring one of the greatest tech companies the world has ever seen. We want to grow Arm and help it become greater. We want more R&D, not less, and we want that work to be done in Cambridge”
Jensen Huang, Nvidia

Speaking on BBC Four’s Today programme, entrepreneur Hermann Hauser described the deal as a “disaster for the UK, Cambridge and Europe”. Hauser believes Nvidia will destroy Arm’s business model, which is based on licensing the Arm chip design to more than 500 chipmakers, most of which are Nvidia competitors. Through the deal, the Arm chip design will be controlled by a single semiconductor firm – as such, Nvidia will have a monopoly on the Arm chip, he warned.

In a blog post written in August, CCS Insight analyst Geoff Blaber said Arm’s independence was critical to its ongoing success. “The moment that control shifts to a rival, this independence is compromised. The value of Arm’s independence is such that an acquisition by Nvidia would be likely to erode Arm’s value.”

But Blaber also claimed Arm was facing growing competition from open source architecture Risc-V. “If its partners believed that Arm’s integrity and independence was compromised, it would accelerate the growth of Risc-V and in the process devalue Arm,” he noted in the blog post.

The AI opportunity

Nvidia has seen its value skyrocket in the past few years, as investors recognise the value of Nvidia graphics processing units (GPUs) being used to power next-generation applications based on machine learning and cryptocurrency mining.

Nvidia has also been pushing forward the use of its autonomous car platform. In December 2019, it introduced its Drive AGX Orin system on a chip based on the Arm processor and has also begun offering an Arm version of its Cuda programming library for developing parallel computing applications.

Cuda enables programmers to write applications that can use co-processors to accelerate algorithms that benefit from being run on highly parallel computing architectures. It has traditionally been used to harness the power of Nvidia GPUs in games programming and AI, where hundreds of GPU cores are tasked with running the same code simultaneously.

With Cuda now supporting Arm, Nvidia has given developers a way to migrate their applications and write new software for Nvidia’s Arm-based systems, such as Drive AGX for autonomous cars.

In June, it announced a partnership with car maker Mercedes. From 2024, next-generation Mercedes-Benz vehicles will include a software-defined computing architecture, built using technology from Nvidia. The aim is to build a hardware platform for autonomous vehicles that can evolve using software as autonomous driving technology improves.

“The value of Arm’s independence is such that an acquisition by Nvidia would be likely to erode Arm’s value”
Geoff Blaber, CCS Insight

In a transcript of Citi’s 2020 Global Technology Virtual Conference posted on the Seeking Alpha financial blogging site, Nvidia’s chief financial officer, Collette Cress, described how the Nvidia hardware platform integrated into a car when it is built could be upgraded with software to enable many different levels of overall autonomous driving during the time it is owned.

Responding to questions over how effective Nvidia will be at maintaining Arm’s independence, Huang said he saw Arm as a cornerstone of the company’s computing strategy. By maintaining Arm’s independence and its business model, Huang believes there is an opportunity to provide Arm licensees with access to Nvidia technology. Potentially, this could mean Nvidia would offer its Cuda programming libraries and development tools for parallel computing to Arm partners, which would hugely expand its software ecosystem.

While it built its reputation on GPUs for gaming, Nvidia sees its future in artificial intelligence. In the age of connected devices, Huang believes AI will move increasingly to the edge, to enable sensors at the edge to make real-time decisions. Such smart sensors require a low-powered, high-performance computing platform, which is what has made the Arm design so successful.

It is no surprise the latest version of its autonomous car platform is based on Arm chips. Given that it has moved beyond GPUs to offer the Cuda development tools on Arm, the acquisition could point to a £31bn bet by Nvidia that it can establish a viable software ecosystem for its technology. 

However, CCS Insight’s Blaber said: “On the face of it, an acquisition of Arm would strengthen Nvidia’s position in silicon for datacentres, the industrial internet of things, and especially smaller client devices. It would give it a high degree of control and scope for customisation. But this is highly theoretical. First and foremost, Arm is a licensing business and offers little real synergy despite the extraordinarily high price tag.”

Read more about edge computing

  • Edge computing is currently one of the most important trends in IT that is likely to complement cloud by supporting new and emerging workloads.
  • In this follow-up guest post Paul Finch, CEO of Harlow-based colocation provider Kao Data, sets out how datacentre designs are having to change to accommodate evolving chip densities.

Read more on Artificial intelligence, automation and robotics

Data Center
Data Management