The British company ARM has taken a significant step in its global expansion strategy within the semiconductor industry. According to Reuters, ARM has hired Rami Sinno, who previously served as the director of AI chip development and design at Amazon. There, he led the creation of the Trainium and Inferentia processors, two critical components in Amazon’s large-scale AI infrastructure.
This move marks Sinno’s return to ARM, where he previously worked from 2014 to 2019, and strengthens the company’s plans to evolve from just designing architectures to developing its own chips.
During his time at Amazon Web Services (AWS), Sinno played a key role in designing AI-optimized chips aimed at competing directly with NVIDIA’s GPUs in terms of cost and performance. The Trainium chips are geared toward large-scale AI model training, while Inferentia is designed for production inference, both critical areas amid the rise of generative AI.
His hiring brings not only technical expertise but also valuable experience managing custom hardware design strategies at one of the world’s largest tech companies.
Traditionally, ARM has focused on designing architectures and instruction sets that are licensed to clients such as Apple, Qualcomm, and NVIDIA. These companies develop their own processors based on ARM technology, paying royalties per chip manufactured. This model has made ARM’s technology ubiquitous; nearly all smartphones worldwide incorporate ARM-based processors, and its influence has expanded into data center servers and IoT devices.
However, in July 2025, ARM announced a strategic pivot: investing part of its profits into developing its own chips and components. Spearheaded by CEO Rene Haas, this shift opens the door for ARM to move beyond licensing revenues and compete directly in the semiconductor market.
The return of Sinno is part of a broader effort to accelerate the design of chiplets and complete systems, aiming to position ARM against both established competitors and emerging startups in the semiconductor space.
SoftBank, ARM’s majority shareholder, views this strategy as a way to capitalize on the booming generative AI market. The demand for chips for training and inference is at unprecedented levels, with the AI accelerator market—dominated by NVIDIA, AMD, and increasingly Google with its TPUs—generating tens of billions of dollars annually.
ARM’s expertise in energy efficiency and modular design could give it a distinctive role in this landscape. Yet, shifting from an IP provider to a leading manufacturer involves significant challenges in investment, timing, and competition.
The return of Rami Sinno signals a pivotal moment for ARM’s strategy. What was once just an idea—becoming a chip developer—begins to materialize now with a strengthened team and leaders who have experience at giants like Amazon, HPE, Intel, and Qualcomm.
ARM aims not only to be the “common language” of mobile and server computing but to become an active player in the race for AI chips, where efficiency, cost, and scalability will determine success.
Frequently Asked Questions (FAQ)
1. Who is Rami Sinno, and what was his role at Amazon?
Rami Sinno was responsible for AI chip development at Amazon Web Services. He led the creation of Trainium and Inferentia, processors designed for large-scale AI training and inference.
2. What sets ARM apart from other chip manufacturers?
ARM traditionally doesn’t build chips. Its model revolves around designing architectures and instruction sets licensed to third parties. Companies like Apple and NVIDIA license these designs to build their processors.
3. Why does ARM want to start designing its own chips?
The goal is to diversify revenue streams, capitalize on the growing AI semiconductor market, and reduce dependence on royalty payments. It also aims to become a direct player in the AI sector.
4. How could this decision impact the semiconductor market?
If ARM successfully develops competitive chips, it could challenge NVIDIA’s dominance in AI and offer more cost-effective and efficient alternatives, affecting data centers and cloud computing.