In a strategic move that underscores its ambition to go beyond its traditional role as a chip architecture provider, Arm Holdings has hired Rami Sinno, Amazon’s director of artificial intelligence (AI) chip development, to lead efforts in building its own silicon products.
The appointment is seen as a major step in Arm’s transformation from a licensing company to a full-fledged chipmaker, as the British semiconductor giant looks to assert itself in a market increasingly defined by AI, cloud computing, and custom silicon.
Rami Sinno was a pivotal figure at Amazon, where he helped spearhead the development of Trainium and Inferentia, the tech giant’s custom-built AI processors. These chips were designed to reduce reliance on Nvidia’s graphics processing units (GPUs), which dominate the AI market, and to provide Amazon Web Services (AWS) customers with more cost-efficient, scalable, and powerful AI solutions.
By bringing Sinno on board, Arm is not just acquiring technical expertise but also insider knowledge of how hyperscalers like Amazon approach chip design for AI workloads. This knowledge could prove invaluable as Arm pushes into new markets that go beyond smartphones and into data centers and generative AI.
Arm’s Traditional Role in the Semiconductor Industry
Until now, Arm has been best known as the architect behind the instruction sets and core designs that power billions of devices worldwide. Companies such as Apple, Qualcomm, and Nvidia license Arm’s intellectual property (IP) and build custom chips around it.
Nearly every smartphone in the world relies on Arm-based processors, and in recent years, Arm’s technology has also made inroads into servers and high-performance computing, challenging incumbents like Intel and AMD.
However, despite its pervasive influence, Arm has not traditionally manufactured its own chips. Instead, it relied on a business model that emphasized licensing IP and collecting royalties on the chips that its customers sold.
That model may now be changing. In July, Arm disclosed plans to reinvest a portion of its profits into building its own chips and other components, signaling a shift toward a more vertically integrated approach.
CEO Rene Haas has openly discussed exploring the creation of chiplet modular, function-specific components that can be combined into complete systems as well as full system-on-chip (SoC) designs.
The rationale is clear: if Arm can create more complete chip designs, it can capture more value from its technology, offer turnkey solutions to customers, and position itself more competitively against rivals who are already vertically integrated.
Sinno is not the first high-profile hire in this effort. Over the past few years, Arm has strategically recruited senior talent from across the semiconductor industry to strengthen its in-house design capabilities.
- Nicolas Dube, formerly of HPE, brings experience in large-scale systems design.
- Steve Halter, who has held engineering roles at Intel and Qualcomm, adds expertise in chip engineering and execution.
By assembling such a team, Arm is laying the groundwork for a transition from being a behind-the-scenes enabler to a direct competitor in certain markets.
AI has become the single most important battleground in the semiconductor industry, with demand for compute power skyrocketing due to the rise of generative AI applications. Nvidia’s GPUs dominate the space, but hyperscalers like Amazon, Microsoft, and Google have increasingly turned to custom silicon to reduce costs and optimize workloads.
Arm’s decision to build its own chips particularly with Sinno’s expertise signals that it wants a bigger piece of this market. With Arm’s energy-efficient designs already favored in mobile and emerging in servers, adding custom AI hardware could make it a serious competitor in data centers.
Moreover, Arm could leverage its ecosystem to provide specialized AI chips optimized for inference and training tasks, offering alternatives to Nvidia’s expensive GPUs.
SoftBank’s Influence and the Business Push
Arm is majority-owned by SoftBank Group, which has been eager to expand the company’s role in the semiconductor landscape and maximize returns following its IPO. Building in-house chips could significantly expand Arm’s revenue streams by reducing its reliance on royalties and increasing its ability to serve new markets like AI and high-performance computing.
This move also aligns with SoftBank’s broader ambitions in AI. By backing Arm’s transition into chipmaking, SoftBank can position itself at the epicenter of the AI hardware revolution, competing not only with Nvidia and AMD but also with hyperscalers that have built proprietary chips.
While the move is ambitious, it is not without risks. Building complete chips is far more capital-intensive and operationally complex than licensing IP. Arm will need to balance this new venture with its core licensing business, ensuring it does not alienate customers who may now see Arm as a competitor.
Additionally, the semiconductor industry is highly competitive, with entrenched players like Nvidia, Intel, and AMD, as well as hyperscalers with deep pockets and in-house design teams. Arm will need to differentiate itself by leveraging its strengths in low-power, high-efficiency designs while rapidly scaling its chipmaking expertise.
Arm’s hiring of Rami Sinno is more than just a talent acquisition; it is a signal of intent. The company is no longer content with simply being the blueprint provider for the world’s chips, it wants to build its own hardware and compete head-on in the AI era.
If successful, Arm could become a formidable force in both mobile and data center markets, bridging the gap between its smartphone dominance and the exploding demand for AI compute.
With SoftBank’s backing, Haas’s vision, and Sinno’s expertise, Arm may be entering a new chapter in its history one where it is not just powering the world’s chips, but creating them outright.
