When you buy through affiliate links in our content, we may earn a commission at no extra cost to you. Learn 
Amazon is intensifying its efforts to loosen Nvidia’s hold on the AI chip market.
The company’s latest AI chip, the Trainium 2 is planned for deployment in its data centers by the end of the year according to Bloomberg.
It signals a major push to offer alternatives to Nvidia’s high-cost GPUs, which currently dominate generative AI workloads but face supply bottlenecks due to soaring demand.
Trainium 2 chip is being spearheaded by Annapurna Labs, an Israeli chip startup that Amazon acquired in 2015 for $350 million. Early customers like Anthropic and Databricks are reportedly testing the chips for their AI workloads.
Amazon’s efforts in this initiative are evident in its increased capital spending in 2024, with an estimated $75 billion allocated to expanding its cloud and chip infrastructure, up from $48.4 billion in 2023.
In a related move, the cloud giant recently announced an additional $4 billion investment in Anthropic, bringing its total investment to $8 billion. This deal includes an agreement for Anthropic to use Amazon’s AI chips more extensively and collaborate on future chip designs
Despite Amazon’s ambitions, dethroning Nvidia’s dominance in the AI processor market is a mammoth task. Nvidia’s second fiscal quarter of 2024 saw a staggering $26.3 billion in revenue from AI data center chips, a figure that rivals AWS’s total earnings for the same period.
Other tech giants, including Google and Microsoft, are also working on custom AI chips to reduce reliance on Nvidia.
Amazon’s ability to catch Nvidia hinges on convincing more customers to adopt its chips while fine-tuning its software and hardware integration. For now, Nvidia remains firmly in the driver’s seat, but Amazon’s investments signal its determination to eventually shake up the AI chip market in the near future.
