The rationale behind NVIDIA Corp.'s acquisition of ARM Ltd. hinges on Nvidia CEO Jensen Huang's vision of a future increasingly dependent on cloud computing and artificial intelligence.

If he is right, the combination of Nvidia's AI capabilities and ARM's processors could make Nvidia as critical to a cloud-based global IT infrastructure as Intel Corp. processors were to the development of the traditional enterprise data-center market, analysts said. Growth for traditional data centers has slowed as cloud computing has become the primary model for high-performance and data-center computing. The challenge for Nvidia will be on finding ways to ramp up ARM technology without alienating its existing business partners.

"They're very different companies and Nvidia is not going to be able to keep ARM independent, which is the whole value of the ARM ecosystem," said Stacy Rasgon, managing director and senior analyst at Sanford C. Bernstein & Co. LLC. "He [Huang] was really committed to driving ARM up into the datacenter. It's clear a lot of value is going to be created there, and that they need the ARM ecosystem to generate the value they're hoping to create."

Unlike Nvidia and other chip companies that sell chips and other devices, ARM makes money by licensing patented designs for chips and electronic components to hardware-makers who pay a licensing fee and royalties on ARM designs they build into their own products. ARM's vast library of intellectual property generated net sales of $457 million during its most recent quarter.

Nvidia, meanwhile, brought in $1.75 billion from sales of its high-performing graphics processors to data-center customers and nearly as much for the sale of desktop versions used by gamers.

Currently, ARM's licensing range is broad enough that most devices with any electronic componentry contain some ARM intellectual property. That works partly because ARM sells only intellectual property. So, for instance, Apple Inc. can use ARM processors as the basis of its iPhone and iPad chips without any concern about Apple's designs being replicated elsewhere, said Linley Gwennap, president and principal analyst at The Linley Group.

Opponents of the ARM deal argue that Nvidia will inevitably harness ARM's enormous library of chip and semiconductor component designs for its own benefit.

Nvidia's Huang has promised not only to preserve the noncompetitive ethos of ARM's highly interdependent ecosystem of customers but also to give ARM's customers access to some of the high-performance graphics and artificial intelligence support that has helped Nvidia dominate the market for data-center GPUs for increasingly common, increasingly demanding AI applications.

Huang sees Nvidia's greatest potential unfolding in a future in which the influence of longtime rival Intel is limited to the market for traditional data-center hardware, which will become less important as the role of cloud computing expands, said Bernstein's Rasgon.

While Intel's decadeslong dominance of the data-center market was once unchallenged, it has struggled with manufacturing in recent years, allowing competitors to make market inroads.

Nvidia's stock rose 164% percent in the 12 months prior to the ARM deal announcement, a run that analysts noted will make the approximately $40 billion ARM deal easier to digest. Intel's stock fell by 6.7% over the same period.

Intel's market entrenchment, including its huge installed base and agreements with server vendors, will keep Intel from crashing hard, but it is almost certain to lose significant market share to Advanced Micro Devices Inc. AMD could be more dangerous to Intel if it were able to make chips in high enough volume to better, Gwennap said. Nvidia has the capacity to meet large-volume demands, but it has been too focused on graphics to compete directly for shares of the CPU market — at least not without ARM's technology.

While Nvidia can make ARM CPUs more competitive with Intel's, Nvidia is more focused on the future. In five to 10 years, Nvidia's Huang envisions a market that is saturated with AI applications that run on Nvidia hardware in a wide range of devices, including phones, edge networks and cloud-based data centers, as he made clear during comments immediately following the Sept. 14 ARM deal announcement.

Intel was successful because its chips were efficient at running traditional applications in traditional data centers on traditional servers, but cloud computing has marginalized all three. Cloud-native software does not have to identify or control server hardware the way traditional applications do, meaning cloud-based applications can be written to run more efficiently and flexibly, Huang noted.

"Because of that evolution ... It is now possible to run on any CPU. It just has to be very energy efficient," Huang said during a Sept. 14 call about the ARM deal announcement.

That has changed the requirements of server processors by enough to allow Amazon Web Services Inc., Google LLC and other cloud service providers — many of which have built AI-application technology using basic ARM processors — to ramp up the power and expand the potential uses of their chips, which can run better than some Intel chips, though only on the cloud platforms for which they were designed, said Paul Teich, principal analyst at market-research firm Liftr Insights.