One of the biggest headlines of this week was the announcement over the weekend that Nvidia would be purchasing Arm for $40 billion from Softbank which purchased the semiconductor IP company back in 2016 for $32 billion. While the rumors of a deal were initially met with a healthy mix of excitement, doubt and speculation, the transaction is now a reality. Over the next few months we will see if the acquisition reaches Day 1.
In a press release, Nvidia CEO Jensen Huang stated, “We are joining arms with Arm to create the leading computing company for the age of AI.” This is a curious statement given Nvidia’s perceived leadership in the AI computing domain with their industry-leading GPUs. Given that Nvidia’s brand and its differentiation in the semiconductor industry is anchored on the GPU, how will the acquisition of Arm impact this narrative?
There is a real risk that Arm, a CPU-centric enterprise, will dilute Nvidia’s unique position in the semiconductor industry as the premiere GPU player. Prior to the rumors of this acquisition, Nvidia had invested a tremendous amount of effort and engineering into developing a wide range of GPU-based platforms and solutions to expand the company’s footprint beyond its traditional graphics processing legacy. Last year, we saw Nvidia introduce its EXG platform for edge AI computing and telco networking. We also saw the continued evolution of the Jetson platform for embedded device applications with the introduction of Nano that furthered Nvidia’s foray into IoT.
While Jensen has described the acquisition of Arm into the Nvidia fold as accretive, the smartphone chip market that Arm dominates has plateaued and continues to experience decline in global shipments which has been reflected in Arm’s top line. The long tail of the IoT market is a low-margin game under constant commodity pricing pressure. In many ways, Nvidia is already well positioned for AI where it will count, in the data center and their new AI edge play for telcos and enterprises, segments that Arm barely has presence in.
On the surface, the union of the portfolios of these two semiconductor powerhouses seems to present significant synergies. With Arm, Nvidia will have the CPU and the heterogenous computing portfolio to take on Intel and AMD and dominate the future of computing. However, there are significant challenges that will make such a hypothesis a risky bet on the prospects for a combined Nvidia-Arm.
First, Nvidia and Arm are fundamentally different businesses. Nvidia is in the business of designing GPU chips and producing graphics cards and other GPU-based systems. Arm is an IP company that licenses their chip technology. This difference in business model will make integration a difficult exercise in maintaining Arm’s ecosystem neutrality which will be critical for preserving the value of Arm. Nvidia will need to design and implement a governance model that will give Arm licensees the assurance that Nvidia’s GPU agenda won’t present a conflict of interest and thus an anti-competitive dynamic down the road.
Second, current geopolitics will make the regulatory approvals difficult to realize given the rapidly souring relationship between the U.S. and China. It is unlikely that Chinese regulators will approve the transaction as Arm-based chips are the mainstay of China’s mobile device industry. Arm landing in the hands of a U.S. company will likely be perceived by the Chinese as a further threat to their technology industry and an uncomfortable consolidation of semiconductor IP under U.S. authority.
A question that has yet to be contemplated is will Nvidia change itself. It would mean taking a “chip neutral” approach of licensing its GPU IP to competing chip makers for free while monetizing their IP through licenses at the OEM level in a fashion similar to Qualcomm’s modem chip licensing model. Such a reinvention of Nvidia’s business model could enable the company to pursue the synergies it aspires to realize while maintain the neutrality the Arm ecosystem depends on for its vitality.
All that being said, if Nvidia can get the regulatory approvals it needs to close the deal, the timing couldn’t be better to challenge Intel’s data center dominance given the semiconductor behemoth’s recently announced delay in their 7nm process
Leonard Lee is the founder and managing director of neXt Curve, a research advisory firm focused on Information and Communication industry and technology research. He has worked as an executive consultant and industry analyst at Gartner, IBM, PwC and EY and has advised leading companies globally on competitive strategy, product and service innovation and business transformation. Follow Leonard on LinkedIn: linkedin.com/in/leonard-lee-nextcurve
“Industry Voices” are opinion columns written by outside contributors—often industry experts or analysts—who are invited to the conversation by Fierce staff. They do not represent the opinions of Fierce.