How Alphabet and Amazon Each Built The ‘Best’ Nvidia Chip Alternatives

Alphabet (NASDAQ: GOOGL | GOOGL Price Prediction) and Amazon (NASDAQ: AMZN) posted Q2 FY2026 results built around the same idea: their custom chips are now real Nvidia alternatives. Google leaned on TPUs powering a 82% Cloud surge. Amazon leaned on Trainium and Graviton driving 37% AWS growth, the fastest in 18 quarters. Two silicon playbooks, two very different customer bases.

TPUs Carry Google Cloud. Trainium Carries AWS.

Google Cloud hit $24.77 billion in Q2, with operating margin jumping to 35.6% from 20.7%. That margin lift is the TPU story in one number: owning the silicon means owning the cost curve. Sundar Pichai told investors Google now offers “the industry’s broadest range of accelerators from Google and NVIDIA”, and notably began recognizing revenue from TPU system sales delivered into customer data centers for the first time.

Amazon went the other way. AWS reached $42.23 billion at a 39.4% operating margin, and Andy Jassy said “our AI and Chips businesses each eclipsed run rates of more than $25 billion” with triple-digit growth. Graviton5 shipped with up to 25% better compute performance than Graviton4, and Trainium capacity is being leased in bulk to Anthropic and OpenAI.

Cost Leader vs. Volume Leader

Google’s edge is efficiency. Pichai flagged that Chrome engineers are “on track to accelerate delivery by eight times” using their own models on their own chips. Amazon’s edge is distribution. Bedrock now hosts Claude Opus 5, GPT-5.6, Gemma 4 and Grok 4.3, and customers spent more on Bedrock in Q2 than in all prior quarters combined. One monetizes silicon through margin. The other monetizes it through rented capacity.

The Next Test Is Free Cash Flow

Both bets are expensive. Alphabet burned $44.92 billion of CapEx in Q2 and posted negative $5.86 billion of free cash flow, with the buyback suspended. Amazon spent $54.21 billion and TTM free cash flow turned negative at $7.6 billion. I want to see TPU external revenue ramp in 2027, and I want Trainium bookings from OpenAI and Anthropic to translate into AWS margin, not just top line.

Why I Lean Toward Google for Silicon Efficiency, Amazon for Silicon Distribution

If you want the cheapest tokens in the industry, Google’s TPU stack looks like the better long-term asset. Cloud margin nearly doubled year over year, and shares rose 5.03% in the past week as investors digested the raise. If you want the widest customer roster leasing custom chips, Amazon wins. AWS jumped 3.9% on earnings day, and the Anthropic and OpenAI Trainium commitments give it a commercial moat Google does not have yet. For investors weighing exposure, the trade-off is efficiency versus distribution, and free cash flow recovery by mid-2027 will be the key checkpoint for both names.

Contact [email protected] for any questions or corrections.