Key Highlights
- Gadi Hutt, who served as director of product and customer engineering at Annapurna Labs, has exited Amazon
- Hutt played a central role in developing Amazon’s Trainium AI chip portfolio
- This marks the second departure of a senior Annapurna executive within a seven-month period
- Rami Sinno moved to Arm (ARM) in August 2025; Rohit Prasad, AGI division head, departed at the close of 2025
- AMZN shares finished Thursday’s session down 2%; year-over-year performance stands at a 3% gain
Amazon (AMZN) finished Thursday’s trading session with a 2% decline.
According to reporting from The Information, Gadi Hutt, who held the position of director of product and customer engineering at Annapurna Labs, has departed Amazon. Hutt served as a prominent figure in the development and promotion of Amazon’s Trainium AI chip series.
Back in 2015, Amazon paid $350 million to acquire Annapurna Labs, a chip design startup based in Israel. Since that acquisition, the division has evolved into a central component of Amazon’s strategy to produce proprietary semiconductors and reduce dependency on external chip vendors.
Hutt’s exit represents the second loss of a senior-level Annapurna executive within the last seven months. Rami Sinno departed in August 2025, subsequently joining Arm Holdings (ARM).
The departure of two critical chip development leaders within such a compressed timeframe creates a noticeable trend.
Beyond the chip engineering group, Rohit Prasad, who served as Amazon’s senior vice president and head scientist focused on artificial general intelligence, also exited the company at the conclusion of 2025.
Multiple Leadership Exits Across AI Operations
The series of departures spans various segments of Amazon’s artificial intelligence infrastructure — ranging from the semiconductor development work at Annapurna to the wider AGI research initiatives previously overseen by Prasad.
Amazon has remained silent on the departures through official channels, and specific motivations behind each individual’s exit remain undisclosed.
Trainium represents Amazon’s proprietary semiconductor architecture built specifically for AI model training operations, and it serves as a cornerstone of the company’s ambition to create cloud computing infrastructure with reduced dependence on Nvidia’s dominant hardware platforms.
Given Hutt’s responsibilities in product development and customer engineering, he maintained direct involvement in determining how Trainium chips were designed, manufactured, and commercialized.
Competitive Landscape for Technical Leadership
The battle for skilled AI semiconductor professionals has intensified throughout the technology sector, with large corporations, emerging startups, and established semiconductor manufacturers all pursuing the same limited pool of engineers and product strategists.
Sinno’s transition to Arm demonstrates the velocity at which leading technical talent can move between competing organizations within this domain.
Amazon’s Annapurna Labs division has served as a foundational element of its AWS cloud platform, where proprietary chips enable the company to deliver distinctive services to business customers.
Trainium semiconductors have been marketed as an economically attractive option compared to Nvidia’s GPU offerings for AI model training applications.
Over the trailing twelve-month period, AMZN stock has advanced 3%, even accounting for Thursday’s 2% pullback triggered by the departure announcement.
