Amazon weighed difficult trade offs as memory costs squeeze devices business, internal document shows

A larger projected loss

The August planning document projected a $612 million gross loss for Amazon’s first-party devices business this year. That is greater than the leadership target of a $438 million loss.

The figures measure gross profit and exclude operating expenses such as research and development, marketing, sales and administration. The document said memory costs were pressuring profitability and requiring trade-offs among sales volumes, average selling prices and gross profit. An Amazon spokesperson declined to comment.

Higher prices, expanding lineup

The company has traditionally used affordable devices to bring customers into its shopping and digital-content ecosystem. Earlier this month, it replaced its budget-friendly Fire tablet lineup with Alexa Tablets powered by Alexa+, priced from $229.99 to $499.99. In August, Amazon also raised prices on products including Echo speakers, Kindle e-readers and Fire TV devices.

AI demand adds cost pressure

Memory chip prices have climbed as data-center investment boosts demand and manufacturers allocate more capacity to high-bandwidth memory for AI servers. Apple, Microsoft, Sony, Dell, HP and Lenovo are also facing component-cost pressure; Apple has raised prices on some Macs and iPads, while Microsoft and Sony have increased console prices.

Amazon expects roughly $220 billion in capital expenditures this year, largely to expand AWS data centers for AI demand. Its planning materials called for updated Alexa+ forecasts, growth through enterprise sales and advertising, and more detailed AWS cost modeling. The company was also seeking to reduce Alexa+ costs by using its own cheaper AI systems more and relying less on Anthropic’s models.

Business Insider previously reported that internal forecasts put Alexa+ AWS cloud costs at roughly $1.7 billion in 2026, nearly three times the prior year’s figure. The separate Moonraker project, intended to give Alexa more advanced AI-agent capabilities, was projected to require more than $100 million in GPU spending this year. Amazon considered delaying or scaling it back as a possible way to ease costs.