Here is the full breakdown of what this shift entails — the rationale, chip approach, partners, timeline, device coverage, and key caveats.

Strategic Rationale: Cost and On-Device AI

The primary drivers for Amazon's custom silicon push are cost reduction and enabling on-device AI. By moving away from off-the-shelf third-party processors (e.g., MediaTek, Qualcomm), Amazon aims to lower per-unit chip costs. This is especially important as its free cash flow narrowed to roughly $1.2 billion in Q1 2026 (about 5% of the prior-year level) amid massive AI infrastructure spending . Custom silicon also allows tight hardware-software co-optimization for running AI models like Alexa+ locally on the device, reducing cloud dependency, latency, and improving privacy .

Chip Design Approach: AZ3, AZ3 Pro, and COT

Amazon has already introduced two custom system-on-chips for consumer devices:

• AZ3 and AZ3 Pro SoCs — Announced in October 2025, these chips are purpose-built for on-device AI inference, running Alexa+ locally rather than in the cloud. The AZ3 powers the Echo Dot Max, while the more powerful AZ3 Pro powers the Echo Studio and Echo Show 8 and 11 . Amazon says the new chips enable better conversation detection and a sensor-fusion platform called Omnisense .
• COT (Customer-Owned Tooling) model — For future consumer processors planned from 2027 onward, Amazon is adopting the same design methodology used for its Trainium AI server chips, giving it in-house control over chip architecture and layout . The goal is not to match Apple Silicon on peak performance but to optimize for cost efficiency and AI-specific workloads at Amazon's scale .

Key Partners: Alchip and Annapurna Labs

• Alchip (Taiwan) — According to well-known industry analyst Ming-Chi Kuo, Amazon has selected Alchip to design the consumer-grade chips for Kindle, Fire TV, Echo, and other gadgets .
• Annapurna Labs — Amazon's own Israeli chip design subsidiary, acquired in 2015, is the internal engine behind Amazon's broader silicon efforts, including the AZ3 family and likely these consumer chips as well .

Timeline: Already Shipping, Broader Rollout in 2027

• Already shipping — Custom AZ3 and AZ3 Pro chips are in current Echo Show 8, Echo Show 11, and certain Fire TV models .
• 2027 onward — A broader transition to in-house processors for Kindle, Echo, Fire TV, Ring, Blink, eero, and other consumer devices begins, with a gradual rollout .
• Target volume — Once fully ramped, Amazon aims to ship roughly 40 million in-house chips per year across its consumer device lines .

Device Coverage: Which Products Are Affected

Key Caveats and What Amazon Has Not Confirmed

It is important to note that analyst reports (Ming-Chi Kuo) and news aggregators are the sole source for the Alchip partnership, the 2027 timeline, and the 40M-unit volume target — Amazon has not officially confirmed those specifics . The consumer chip strategy is framed primarily as a cost-saving measure to free up capital for AI infrastructure investment, not as a bid to create a high-performance "Apple Silicon" competitor . Amazon still uses third-party chips from companies like Qualcomm for some products .

Here is the full breakdown of what this shift entails — the rationale, chip approach, partners, timeline, device coverage, and key caveats.

Strategic Rationale: Cost and On-Device AI

The primary drivers for Amazon's custom silicon push are cost reduction and enabling on-device AI. By moving away from off-the-shelf third-party processors (e.g., MediaTek, Qualcomm), Amazon aims to lower per-unit chip costs. This is especially important as its free cash flow narrowed to roughly $1.2 billion in Q1 2026 (about 5% of the prior-year level) amid massive AI infrastructure spending . Custom silicon also allows tight hardware-software co-optimization for running AI models like Alexa+ locally on the device, reducing cloud dependency, latency, and improving privacy .

Chip Design Approach: AZ3, AZ3 Pro, and COT

Amazon has already introduced two custom system-on-chips for consumer devices:

• AZ3 and AZ3 Pro SoCs — Announced in October 2025, these chips are purpose-built for on-device AI inference, running Alexa+ locally rather than in the cloud. The AZ3 powers the Echo Dot Max, while the more powerful AZ3 Pro powers the Echo Studio and Echo Show 8 and 11 . Amazon says the new chips enable better conversation detection and a sensor-fusion platform called Omnisense .
• COT (Customer-Owned Tooling) model — For future consumer processors planned from 2027 onward, Amazon is adopting the same design methodology used for its Trainium AI server chips, giving it in-house control over chip architecture and layout . The goal is not to match Apple Silicon on peak performance but to optimize for cost efficiency and AI-specific workloads at Amazon's scale .

Key Partners: Alchip and Annapurna Labs

• Alchip (Taiwan) — According to well-known industry analyst Ming-Chi Kuo, Amazon has selected Alchip to design the consumer-grade chips for Kindle, Fire TV, Echo, and other gadgets .
• Annapurna Labs — Amazon's own Israeli chip design subsidiary, acquired in 2015, is the internal engine behind Amazon's broader silicon efforts, including the AZ3 family and likely these consumer chips as well .

Timeline: Already Shipping, Broader Rollout in 2027

• Already shipping — Custom AZ3 and AZ3 Pro chips are in current Echo Show 8, Echo Show 11, and certain Fire TV models .
• 2027 onward — A broader transition to in-house processors for Kindle, Echo, Fire TV, Ring, Blink, eero, and other consumer devices begins, with a gradual rollout .
• Target volume — Once fully ramped, Amazon aims to ship roughly 40 million in-house chips per year across its consumer device lines .

Device Coverage: Which Products Are Affected

Key Caveats and What Amazon Has Not Confirmed

It is important to note that analyst reports (Ming-Chi Kuo) and news aggregators are the sole source for the Alchip partnership, the 2027 timeline, and the 40M-unit volume target — Amazon has not officially confirmed those specifics . The consumer chip strategy is framed primarily as a cost-saving measure to free up capital for AI infrastructure investment, not as a bid to create a high-performance "Apple Silicon" competitor . Amazon still uses third-party chips from companies like Qualcomm for some products .