06-reference

not boring base power chapter 3

2026-08-12·reference·source: Not Boring·by Packy McCormick
not-boringbase-powerenergybatteriescompany-profilepower-cyclecapital-cycle

"Base Power Company: Chapter 3" — @packyM

Why this is in the vault

Third installment in Packy's ongoing company-profile series on Base Power — a home-battery "gentailer" that just raised $1B at a $13B post-money valuation and is explicitly pivoting into AI-data-center power infrastructure, making it a live, named-company data point for RDCO's active power-cycle-v1 investing thesis.

The core argument

Base Power, a ~3-year-old Austin startup, closed a $1B Series D (Ribbit, Valor, Addition, JPMorgan's Strategic Investment Group) at a $12B pre-money / $13B post-money valuation — up from $4B roughly nine months earlier. Base's model: it's a Texas "gentailer" — a Retail Electric Provider that also owns generation — installing home battery systems that give homeowners whole-home backup while letting Base (and utility partners like Austin Energy) dispatch the fleet during demand spikes.

Packy's frame is the "Capability Cannon": Base has spent three chapters accumulating deployment, manufacturing, brand, balance-sheet, and policy muscle in one narrow lane (Texas home batteries), and is now repointing that accumulated capability at adjacent markets — new states (Illinois joins Texas), in-house Gen 2 battery manufacturing (accelerated by 2025 tariff pressure, replacing off-the-shelf Gen 1 hardware), and — new in this chapter — AI data-center power, described as "bringing the power to the compute, and... bringing the compute to the power." CEO Zach Dell frames the ambition against the largest energy companies in the world (Saudi Aramco $1.7T, Exxon $650B), summarized in his own words: "we should be able to build the largest energy technology company ever." Cited growth metrics from an investor update: installs crossing 400/month, 30% MoM growth in closed customers, and Austin Energy advancing Base through a formal utility-selection process.

Series continuity: Chapter 1 covered why Base chose the REP/gentailer structure (grounded in a 2023 Casey Handmer essay on grid batteries); Chapter 2 (~15 months prior) covered the single-state Texas pilot and an interview with Head of Hardware Dino Sasaridis about deliberately delaying in-house manufacturing (crawl-walk-run) before committing to Gen 2 production. This chapter is the payoff of that patience: manufacturing now in-house, footprint expanding, and the AI-power pivot layered on top.

Mapping against Ray Data Co

The concrete connection: RDCO's active power-cycle-v1 investing thesis (~/rdco-vault/01-projects/investing/theses/2026-05-17-power-cycle-v1.md) names power generation and transmission — not chips, not capex willingness — as the binding constraint on AI buildout, and tracks demand-side anchors like hyperscaler PPA cadence and interconnection-queue backlog. Base Power's own pivot line in this essay — explicitly building to "bring the power to the compute" — is a named-company, real-time corroboration of that exact bottleneck thesis, independent of and predating this essay. Base itself is private and outside the thesis's locked public-market ticker universe (IPPs, nuclear, uranium), so this isn't a portfolio-overlap signal — it's evidence the underlying demand thesis (power as the scarce layer under AI infrastructure) is being acted on by venture-scale capital at a $13B valuation, which is exactly the kind of corroborating signal the thesis's disqualifying-conditions framework is built to watch for (the opposite direction: if power bottleneck plays like Base stopped attracting this kind of capital, that would read as a bearish flip).

Secondary, weaker mapping: the "Capability Cannon" concept (accumulated organizational skill in one lane, repointed at adjacent markets) rhymes with how RDCO's own design system is structured — the umbrella system in ~/.claude/skills/ray-data-co-design/SKILL.md explicitly inherits into each sub-bet's design skill (Sanity Check, MAC, Squarely) rather than each bet rebuilding capability from scratch. It's a real structural echo but not a decision-grade insight on its own; noting it for pattern-tracking rather than claiming it as load-bearing.

No connection to the chip-fab/memory capital-cycle thesis specifically — Base Power is batteries/grid, not semiconductors, and prior WDoO coverage (2026-08-07 entry) already flagged that Base Power items reinforce the general physical-AI/hard-energy optimism theme without adding new RDCO-specific signal beyond that pattern. This note doesn't force that connection.

⚠️ Sponsorship

No paid ad block in this issue — this is a straight essay, not a sponsored placement. But Packy discloses, plainly and early in the piece, that Base Power "is certainly my aim" for the generational-company format and is a company he has "both invested in and plan[s] to keep writing about through and beyond IPO." That's an unhedged, self-disclosed Not Boring Capital investment in the subject of the essay — the strongest form of the structural-sponsor pattern the README's Not Boring gotchas describe (co-written founder pitches without a paid block), except here it's solo-authored and the conflict is explicit rather than inferred. Read the valuation framing, the "largest energy technology company ever" quote, and the AI-data-center pivot as an investor's bull case in essay form, not neutral reporting — the underlying facts (raise size, valuation, growth metrics) are still citable, but the interpretive frame is not disinterested.

Related