Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more

Amazon acquired Israeli chip designer Annapurna Labs in 2015. The often-cited $350 million price comes from contemporaneous press coverage, not an acquisition-specific figure confirmed in the Amazon and SEC passages cited here. The acquisition’s better-documented legacy is AWS custom silicon: Nitro infrastructure, Graviton processors, and machine-learning chips Inferentia and Trainium.

Did Amazon buy Annapurna Labs?

Yes. Amazon acquired Annapurna Labs in 2015. AWS Chief Evangelist Jeff Barr described the timing and the relationship in a 2018 post introducing A1 EC2 instances: “We acquired Annapurna Labs in 2015 after working with them on the first version of the AWS Nitro System.” AWS’s A1 announcement places the deal in the development of cloud infrastructure, rather than consumer electronics.

Annapurna was a specialist microelectronics company. Amazon later credited its team with helping develop custom silicon for AWS services and infrastructure. The acquisition did not turn Annapurna’s chips into retail products for ordinary computer buyers; the technologies described by Amazon are used within AWS.

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

Was the purchase price really $350 million?

$350 million is a contemporaneously reported price, not a deal value independently confirmed in the official Amazon and SEC passages cited here. The distinction matters: the figure is useful context, but it should be described as reported rather than as an officially disclosed purchase price.

Amazon’s 2015 Form 10-K reports an aggregate purchase price of $690 million for certain companies acquired that year. The cited filing passage does not name Annapurna Labs or break out a price for it, so it cannot verify the $350 million figure. Amazon’s 2015 Form 10-K

What did Annapurna’s team build for AWS?

Amazon’s account of the team’s work traces a progression from infrastructure offload to general-purpose cloud computing and then machine learning. These product lines serve different workloads; they are not interchangeable consumer chips.

Technology AWS role
Nitro System Supports AWS networking, storage, and security. AWS says its team worked with Annapurna on the first version before the acquisition.
Graviton Arm-based processors for general cloud computing, including the A1 EC2 instance family introduced in 2018.
Inferentia Machine-learning inference.
Trainium AI model training and inference.

Amazon’s retrospective describes Annapurna’s evolution within its custom-silicon effort, from the Nitro work to Graviton and later machine-learning chips. Amazon’s account of its custom-chip work and AWS’s 2018 A1 announcement provide the company’s descriptions of this lineage.

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Support on Ko-Fi

Why the acquisition mattered to AWS

Cloud providers can design chips and systems around the services they operate, rather than relying only on general-purpose processors and components. In AWS’s account, Nitro handles infrastructure functions such as networking, storage, and security; Graviton supplies processors for general cloud workloads; and Inferentia and Trainium address machine-learning workloads. Together, these examples show why Amazon valued a specialist silicon team: custom hardware could be integrated into AWS’s own infrastructure and instance offerings.

Amazon said in 2026 that its custom-chip business had recently exceeded a $25 billion annual revenue run rate. That is Amazon’s company-reported figure, not an independently verified measure, and it refers to the broader custom-chip business—not Annapurna Labs alone. Amazon’s 2026 custom-chip update

Best Value
Samsung 16” Galaxy Book5 Pro 360 Copilot+ PC, AI Business Laptop, Windows 11 Pro, Intel Core Ultra 7 Processor 258V, 3K AMOLED Touchscreen, 32GB / 1TB, 120HZ, 2025 Model NP964QHA-KG2US, Gray
  • POWER FOR YOUR MOST PRODUCTIVE DAYS: Galaxy Book5 Pro 360 models come with a new Intel Core Ultra 7 processor 256V with a higher-performance Intel ARC graphics & newly added AI neural processing unit (NPU) that reaches as high as 47 TOPS to achieve Copilot+ AI.
  • POWERFUL. DURABLE. LIGHT. AMAZINGLY SLIM: Galaxy Book5 Pro 360 is the epitome of portability and the lightest in our all-new Galaxy Book5 Series but very durable carved from CNC Aluminum; Thickness: 12.7mm, Weight: 3.72 lbs.
  • EVERYTHING UNDER MULTI-CONTROL: Navigate your phone, Galaxy Tab, Samsung TVs, and Smart Monitors using your laptop's touchpad and keyboard.¹
  • PHONE LINK: Sync your Galaxy phone to access AI features, call, text, and more— all without touching your phone.²
  • TRANSCRIPT ASSIST: Turn recorded lectures and meetings into written notes, then translate and summarize quickly with Galaxy AI.³

Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.