T-Mobile's fallow capacity model shapes FWA and spectrum strategy in 2026 Q2

How T-Mobile frames traffic growth, upper C-band/2.0/2.7, and FWA under a “fallow capacity” model

1) Does T-Mobile expect a step-change in traffic from AI workloads like AT&T?

Based on the provided transcript, T-Mobile has not “seen it as yet,” but is looking forward to it. Management explicitly says: “We haven't seen it as yet, but we look forward to seeing it because our network is more prepared than anyone else.” ¹

They do not give a quantified forecast in the excerpts for how much “AI workloads” will add to traffic, nor do they confirm that the AI-driven growth magnitude will match another carrier’s expectations. Instead, they frame the situation as an opportunity tied to network evolution (2.7/C-band-related capacity and future 6G capabilities), while remaining “too early to tell” on the exact opportunity shape in 2028. ²³

Net:

• AI-driven traffic growth: Not yet observed in their operating view (“haven't seen it as yet”). ¹
• Readiness: They believe their network is better positioned to handle emerging demand. ¹
• Quantification: No specific magnitude is provided in these excerpts. ¹²

2) What is the “fallow capacity model,” and is it actually “fallow” by design?

T-Mobile’s answer is very direct: it’s a complete fallow capacity model. ⁴

They describe the mechanics:

• They forecast wireless traffic at the hexbin level across 36 million hexbins. ⁴
• They then apply a market-share cap on FWA demand (i.e., limit how much of that forecast they expect to win on FWA). ⁴
• The remaining capacity is “fallow capacity,” and everything is done at peak hour. ⁴

They also reinforce the utilization/capacity relationship:

• FWA consumes a fair amount of traffic today, but not the same thing as consuming capacity, because capacity is “several times multiple of the traffic” they currently hold. ⁴
• They further clarify utilization is “completely by design” and that capacity is way in excess of current utilization. ⁵
• They call out that the 15 million subscriber framework included “conservative rational market share caps,” reinforcing the disciplined capacity planning behind their targets. ⁶

Net:

• The model is explicitly designed to leave capacity unused under a capped FWA adoption assumption. ⁴⁵⁶
• FWA demand is treated as traffic load rather than a full capacity “fill,” given their stated peak-hour surplus. ⁴⁵

3) Is upper C-band (and related auctions) intended for FWA or for AI workloads?

In the excerpted discussion, management treats spectrum (including upper C-band and other bands) primarily as part of a capacity and technology leadership platform, not as a single-purpose tool “specifically for AI” or “specifically for FWA.”

3.1 What they say about spectrum/auctions and network evolution

They characterize spectrum auctions as including:

• Upper C-band as a “big announcement,” plus
• 2.7 GHz (mentioned as part of the Chairman’s comments). ⁴

They then argue the broader strategic posture:

• Due to rapid wireless evolution versus other broadband technologies, they are “double down[ing] on wireless tech” and building out network superiority for 5G and 6G. ⁴

3.2 How they describe the role of FWA in that future

They explicitly connect FWA to the capacity planning framework:

• “Yes, FWA does consume a fair amount of our capacity” (as discussed), but they emphasize the “numerator/denominator” point: it consumes traffic, while capacity is several times higher than current traffic. ⁴
• They say they remain under the fallow capacity model even while discussing incremental spectrum. ⁵

They also add that:

• They are not going to purchase spectrum “strictly for fixed wireless” while still staying within the fallow-capacity construct. ⁵

3.3 What they say about upper C-band sequencing and capabilities (including AI enablers)

For 2.7 specifically:

• They mention an opportunity in “'27 and '28” and note 2.7 in the 2028 timeline as part of cementing network leadership. ²
• They frame future differentiation as linked to 6G-related opportunities including “edge AI and physical AI.” ⁷
• However, they also emphasize that it is “too early to tell” what those opportunities will look like next year and in 2028. ²

Net:

• Spectrum strategy is presented as network leadership and future capability enablement (including 6G/edge/physical AI), rather than purely FWA. ⁴⁷
• FWA remains important, but T-Mobile positions it as operating within (and not breaking) a fallow capacity plan, not as the sole destination for additional upper C-band/2.7 capacity. ⁴⁵

4) How FWA “fits” into the future under the fallow model

T-Mobile’s view of FWA under a fallow-capacity regime can be summarized as:

4.1 FWA is capped and capacity surplus is intentional

• They forecast and then cap FWA market share; the rest of capacity is planned as fallow. ⁴
• They reiterate utilization is designed by the model and capacity is way in excess of current use. ⁵

4.2 Incremental spectrum enlarges both “product quality” and “fallow capacity”

They state that guidance for 15 million customers was built excluding certain factors, and then:

• New spectrum “coming online” makes them more positive because guidance didn’t build any of that in. ³
• That spectrum “actually creates more fallow capacity as well.” ³

This is the core reconciliation to the user’s “fallow capacity strategy” concern:

• If FWA traffic grows, the planning response is not to assume capacity scarcity; management suggests capacity additions expand the surplus buffer that underwrites FWA growth. ³⁴

4.3 How competing technologies (fiber, LEO) affect the FWA stance

While not directly “upper C-band vs AI,” T-Mobile does discuss complementary behavior that reinforces flexibility:

• They argue fiber and FWA are complementary under their model: winning fiber frees up fallow capacity in those regions, which can be sold elsewhere because mobile cell radiuses are large. ⁸

They also address LEO vs FWA at a high level by emphasizing:

• Product superiority driven by network superiority is “the heart” of their success. ⁹
• They emphasize speed/product delivery over WiFi as “fiber-like” (a differentiator), supporting their confidence without abandoning the capacity framework. ⁹

4.4 FWA runway and demand composition

They provide some indicative demand-runway evidence:

• They state 2/3 of customers are in the top 100 markets and shares are lowest in top 100—implying runway ahead for FWA. ¹⁰

They also maintain that strong wireless broadband performance is visible operationally:

• They call out “another very strong quarter” for fixed wireless broadband with total broadband” in the “upper 400,000 range” and “strong ARPUs.” ¹¹

Net:

• FWA is treated as a bounded-growth, market-capped use of a much larger planned capacity pool. ⁴⁶
• Additional spectrum is viewed as increasing both capacity quality and the size of the surplus buffer (“more fallow capacity”). ³
• The strategic emphasis remains network superiority as the driver, with AI/6G capabilities viewed as potential future demand/technology accelerants. ⁹⁷

5) Direct answer to the user’s framing: capacity utilization and “fallowness”

You asked whether T-Mobile’s business is “consuming most of the capacity” and is “extensively a fallow capacity strategy.”

From management’s words in the excerpts:

• FWA consumes a fair amount of traffic today, but T-Mobile insists it does not consume the capacity in the same way, because capacity is several times multiple of current traffic. ⁴
• They state utilization is completely by design and capacity is way in excess of current utilization. ⁵
• The “fallow capacity” is not accidental—it is created through peak-hour forecasting and market-share caps at a granular hexbin level, with the remainder explicitly defined as fallow capacity. ⁴

So, within the information provided here, their position is that the network is not operating at “most of capacity”—it’s operating with intentional surplus to support demand growth (including FWA growth) while protecting economics and scalability. ⁴⁵

---

Bottom line: T-Mobile does not claim to have already observed the AI-driven traffic step-change, but management believes their network is prepared. ¹ They frame future auctions/spectrum (including upper C-band and 2.7) as part of a broader wireless/6G leadership strategy, while keeping FWA within a complete fallow capacity model that caps FWA market share and preserves capacity surplus—indeed, new spectrum they discuss can increase fallow capacity further. ⁴⁵³