Alphabet is now selling its TPU hardware directly to select outside customers, and the first systems have reached other companies' data centers. The company expects most of the sales tied to these agreements to arrive in 2027.
Alphabet is selling its Tensor Processing Units directly to select outside customers, a shift from renting the chips through Google Cloud. The first systems have already landed in other companies' data centers.
From rental model to hardware vendor
Alphabet first disclosed the plan in late April 2026, during its first-quarter earnings call. Until then, the chips mostly powered Google's own operations, and outsiders had to rent capacity through Google Cloud.
On the second-quarter call in July, CFO Anat Ashkenazi confirmed that the first TPU systems reached external customer data centers during that quarter. Those deliveries allowed Alphabet to recognize its first revenue from direct TPU sales.
However, the company expects most of the sales tied to these agreements to arrive in 2027. Alphabet has also signaled that quarterly TPU hardware results are expected to swing around.
Who is buying, and how big it could get
Demand comes from AI labs, capital markets firms and high-performance computing users. Meta began running TPUs for production workloads in its own data centers as of October 2026.
Oracle is also on board. The two companies announced a partnership that will make TPUs available on Oracle Cloud Infrastructure starting in calendar 2027, making Oracle the first hyperscaler outside of Google to offer the chips.
Analyst projections put external TPU revenue at approximately $8 billion in 2026. For 2027, forecasts range from $51 billion to more than $75 billion.
The sales show up inside Google Cloud's backlog, which reached $514 billion by the end of the second quarter and now includes these TPU commitments. Google Cloud revenue climbed 82% year over year to $24.8 billion in Q2 2026, a figure that does not reflect the impact of TPU sales.
What it means for Nvidia and the cloud market
Alphabet's stated reasoning is twofold: selling hardware directly expands its addressable market, and the business is meant to help fund and support development of its next-generation silicon.
The risks are real. Alphabet is still working on making TPUs more compatible with the software tools developers already use, and that gap is one of the main reasons customers default to Nvidia.
Investors will want to see whether Alphabet starts disclosing TPU revenue directly, how the 2027 delivery schedule plays out, and whether more cloud providers follow Oracle's lead.
Source: Crypto Briefing
Trading involves risk.
Risk Disclaimer
