NVIDIA, the Silicon Valley-based chipmaker, has agreed to buy the UK’s Arm for $40 billion — a record semiconductor acquisition that will create a global powerhouse with a sweeping CPU and GPU portfolio. (The computer chips that respectively, power “general” workloads, and accelerate applications like graphics, supercomputing and AI by dividing tasks among multiple processors).
NVIDIA will “continue Arm’s open-licensing model and customer neutrality and expand Arm’s IP licensing portfolio with NVIDIA technology” it pledged. Cambridge will remain Arm’s HQ. Customers can expect beefed up, large core chips for data centres and more powerful mobile GPUs as portfolios merge.
The acquisition, if approved by competition regulators over the next 18 months, is likely to result in a rise in use of (rival) open source, royalty-free RISC-V designs by companies concerned at a further narrowing of the global semiconductor technology market; particularly in China, which is already ramping up efforts to break dependence on Arm IP — now a de facto standard for mobile devices.
“To attract researchers and scientists from the U.K. and around the world to conduct groundbreaking work, NVIDIA will build a state-of-the-art AI supercomputer [in Cambridge], powered by Arm CPUs”, the company added, pledging to grow the site and set up a technology incubator for good measure.
Arm’s licensees have shipped over 180 billion chips since its founding in 1990. The company offers a wide range of CPU and GPU designs; traditionally used most heavily in smaller devices, but increasingly being used to power the IoT, laptops and indeed, cloud workloads; its chips touch nearly every market segment.
The deal comes as Apple, Amazon and Microsoft all ramp up their use of Arm chip — which have significantly different Instruction Set Architecture (ISA) to the widely used x86 ISA in Intel’s chips.
Critics say that Arm’s documentation remains inconsistent, with few go-to hubs as handy as Intel’s Intrinsics Guide website: an interactive reference tool that allows developers to work with Intel chips without the need to write assembly code.
But the company’s IP is developing fast, its architecture hugely flexible, and the move comes amid a broader global push by a range of major semiconductor buyers to diversify their supply chains and offer Arm-powered workloads.
“AI is the most powerful technology force of our time and has launched a new wave of computing,” said Jensen Huang, founder and CEO of NVIDIA. “In the years ahead, trillions of computers running AI will create a new internet-of-things that is thousands of times larger than today’s internet-of-people. Our combination will create a company fabulously positioned for the age of AI.
“Simon Segars and his team at Arm have built an extraordinary company that is contributing to nearly every technology market in the world. Uniting NVIDIA’s AI computing capabilities with the vast ecosystem of Arm’s CPU, we can advance computing from the cloud, smartphones, PCs, self-driving cars and robotics, to edge IoT, and expand AI computing to every corner of the globe.
“NVIDIA is the perfect partner for Arm,” added Masayoshi Son, chairman and CEO of SBG, in a canned statement adding: “This is a compelling combination that projects Arm, Cambridge and the U.K. to the forefront of some of the most exciting technological innovations of our time and is why SoftBank is excited to invest in Arm’s long-term success as a major shareholder in NVIDIA.”
Not everyone is a fan of the proposed deal: Arm founder Dr Hermann Hauser is among those to have spoken out vociferously against the deal today, warning it could “destroy the very basis of ARM’s business model which is to be the Switzerland of the semiconductor industry dealing in an even-handed way with its over 500 licensees. Most of them are Nvidia’s competitors…”
Competition regulators may, indeed, take a very close look at the deal, which further narrows the global GPU market. The boards of Arm, NVIDIA and SoftBank, meanwhile, have signed off on the deal, and completion of the transaction is expected to take place in approximately 18 months.
Geoff Blaber, VP Research, Americas, CCS Insight, was also unimpressed.
He said: “Acquisition by Nvidia would be detrimental to Arm and its ecosystem. Independence is critical to the ongoing success of Arm and once that is compromised, its value will start to erode. It would accelerate the growth of RISC-V as an open-source alternative”
“This will rightly face huge opposition, most notably from Arm licensees who have collectively shipped an average of 22 billion chips annually over the last three years. A huge diversity of businesses from Apple to Qualcomm are dependent on Arm and will be motivated to unite in opposition. Nvidia has a mountain to climb in securing regulatory clearance. This process will take months if not years with a high chance of failure. This process would be damaging to all parties and the uncertainty alone would hurt Arm regardless of the outcome”.
“On the face of it an acquisition of Arm would strengthen Nvidia’s position in silicon for data centres, the industrial Internet of things, and especially smaller client devices. It would give it a high degree of control and scope for customization. But this is highly theoretical. First and foremost Arm is a licensing business and offers little real synergy despite the extraordinarily high price tag”.