🫡 Happy Tuesday gang. Last week I promised you Dan’s full thinking on ramping launch cadence and capacity, which is all folks in our circles seem to be talking about right now.
Today he forcefully delivers, with an Antimemo arguing America “should stop buying rockets and start buying orbit.”
Riding shotgun to Dan’s tour de force on space transport, we have:
🏗️ A guest dispatch from Closeshoring CEO and friend-of-PA Alan Simon on how to actually sell the datacenter buildout
📊 Companion reading re: Dan’s essay: your verdict on a 1,000-launch target… results are in
🛠️ The venture lesson from The Home Depot
🐕 Good news for Man's Best Friend; Around the Horn on Wall Street; and ATX Happenings
📸 Public works of art, c/o California's water bureaucracy (yes, really)
Forwarded this? Join Per Aspera today.
DG here. In the early 1970s, as the Space Shuttle was taking shape, my old boss at TRW, Gerry Elverum, proposed a rocket engine that dispensed with turbopumps altogether: a pressure-fed design simple enough to be built in a pipe factory, for ~$20,000, at a time when the Shuttle's main engines ran to millions apiece and depended on turbomachinery so troublesome that many years later I had to direct its redesign as administrator. NASA declined Gerry’s design. The objection, verbatim, was that "the design wasn't sufficiently advanced." It was an absolutely ridiculous response!
Today I’m sharing an Antimemo about why that instinct still runs American space policy: America Is Buying Rockets When It Should Be Buying Orbit. We still shop for launch the way we “shopped” for the Shuttle — by writing a magnum opus of requirements, anointing a contractor, and assuming the aging government-operated range can accommodate you.
Fifty years of this habit has left us with one company that can actually fly, and a country waiting in line for access to its own sky.
Last year the U.S. flew 181 orbital launch attempts, up from 145 the year before. The number was received as a triumph. But when you size it against inbound demand, it should be a warning. The White House would agree: it wants 1,000+ American launches a year by 2030. Once you fold in the constellations the Department of War wants, the intelligence community's plans, and hyperscalers eyeing orbital compute, I believe 5,000 to 10,000 is plausible within a decade.
Meanwhile, the vehicle flying nine of every ten American launches is being wound down as SpaceX retools for Starship, and the Cape is, by Space Force leadership's own admission, "very, very close to out of pads."
This isn't penciling. And we’re acting as though time is on our side. On this one, our pants are on fire! 👇️ Give it a read + write back with your thoughts.
A Hot Take: Datacenters Are Good for America
Alan Simon here. Last week: Per Aspera published my chart comparing data centers' projected '28 footprint to other critical American infrastructure (golf courses included 🙂).
My latest here isn't about whether data centers are good or evil (clearly, it’s all relative 👆) — but that earning support for data centers, and for much of the other infrastructure we're all currently building out, requires development models that benefit folks who live nearby.
Which require: coordination with local communities. Musings on how we can have both data centers and thriving communities! (Did you read the NL where Ryan mentioned that PA is generally net optimistic?!)
Site — site smaller, modular data centers away from homes — brownfields, established industrial parks, and the like to minimize disruption and let infrastructure grow in manageable stages.
Cluster — cluster computing with adv manufacturing and other job creating businesses. I've heard that industrial districts pairing data centers with precision mfg, robotics, materials processing, eng services, and technical training centers (aka. Tribal Knowledge) build stronger regional economies.
Design - give communities a real say in exterior design. These buildings shape the landscape for decades, and a windowless box offers little to be proud of. Beauty belongs to the voice (or at least the most aesthetic voices) of the people!
Lower bills — developers should cover the burden their projects impose, protect existing ratepayers, and add supply where needed. In local terms: Local electricity bills should go down, not up.
This is the coordination required to chain the AI buildout directly to American industrial revival (computing → energy → production → design → workforce, which all reinforce one another when planned as pieces of a successful local economy).
Today, one of the harder parts of being an American industrialist is earning the support of the community. So this is a call to the builders: that support comes from intentionally planning and building so the communities around our work become more prosperous (which is I would say is the definition of a patriot builder).
THE VERDICT ON 1,000 LAUNCHES. Last week (#062), we talked about NSPM-17, the White House's first rewrite of national space transportation policy in a decade-plus, which directs U.S. ranges to support 1,000+ launches and reentries a year by 2030. (The YTD count: 176.) Given the distance between the current manifest and this mandate, we wanted to know where y’all think this could actually land. So we asked: how many launches will America hit in 2030?
Only ~14% of you are believers, taking the White House at its word. The midpoint settled around 500 — call it 3x today's cadence, still a country mile short of our target and still a wildly bullish number by historical standards. Hardly a bear in the room, in other words: 77% of you have America doubling its all-time record pace, or better, inside four years. The disagreement is all about slope, and implicitly, whether we’ve got what it takes to accommodate the high-rate cadence that we need. One of you zeroed in on monopoly math:
The US only has one REAL player. (The WORLD only has one.) It's a monopoly position… If you take SpaceX out of the equation (because now it's a data/AI company that happens to launch rockets as its side hustle), the math gets bad in a hurry: 28 US launches, that's just 11 above the failed-state-but-doesn't-know-it-yet of Russia's 17 launches… So my prediction is that SpaceX will drop (more Starships, but fewer Dragons = more tonnage to orbit). And that the also-rans of our industry will ramp up. So I'm looking at 350-400 range.
To put things in context: for the optimists to be right — for us to hit 1,000 a year — American ranges would be turning nearly three launches a day.
Where do we land? Well, count us as hopeful optimists but stubborn realists. As we lay out in today’s main essay, it will be a herculean task to take near-airline tempo and apply it to orbital flight, with pads, ranges, and airspace deconfliction scaled to match. But demand won’t be an issue. The problem is our scarcest inputs: concrete, coastline, and the will to get out of our own way.
Source: FactSet
Venture capitalists are supposed to be good at recognizing extraordinary companies early. But even the best investors have blind spots — and sometimes the biggest misses come from passing on a business that doesn’t pattern-match. As Ian Rountree recently noted on X, his late uncle and mentor Dick Kramlich — co-founder of NEA and an early Apple backer — once passed on Home Depot’s seed round because he didn’t consider it a “venture-style” business. A fun fact, and a costly one in hindsight.
The rationale was entirely logical: Home Depot was a retail company, not a software startup or technology platform.
NEA weighed the deal at $10M pre-money, brought in two retail experts, and both advised against it (they hated the plan to jump to California before densifying Atlanta).
According to Chuck Newhall, another NEA cofounder: “Sometimes experts don’t know everything they’re supposed to.”
All of this was prompted by the latest Acquired podcast — which you really ought to check out. The Home Depot is one of the most successful stories in American business: founded in Atlanta in 1978 by Bernie Marcus, Arthur Blank, Ron Brill, and Pat Farrah, the company opened its first two stores a year later and IPO’d in 1981. Per the Acquired team, The Home Depot went on to become the highest-returning U.S. stock of the last 45 years: ~24% annualized with dividends reinvested, beating Apple, Microsoft, and Berkshire. ($1,000 invested at the ‘81 IPO would be ~$16M today.)
Which reminds us of a thread from last week’s Hard Reflections interview with Overmatch’s Jordan Blashek — and of something PA cofounder and our neighborhood VC Jeff Crusey will tell you: “be contrarian” is the most consensus advice in the business. Everyone dispenses it. Actually practicing it is a different thing altogether, but it’s where all of the edge lives. A lot of times, the opportunity is hidden in an unglamorous category, an overlooked operating model, or a business that ‘experts’ believe is too unconventional to pull off.
So… who’s building the next Home Depot? Holler @ Jeff!
WHAT’S UP BIG DOG??? // Loyal, the SF biotech developing longevity drugs for dogs, has just landed its third FDA acceptance, with LOY-003, a daily pill designed to help big dogs live longer, healthier lives, earning the agency’s “reasonable expectation of effectiveness” nod last week. When Celine Halioua (a UT Austin neuroscience grad 🤘 who dropped out of an Oxford PhD to chase this problem, started the company as a solo founder, and is someone we’re huge fans of, BTW) set out on this mission, no FDA pathway existed for any lifespan-extension drug. Loyal got one invented: in 2023, the FDA accepted a longer-lived dog as an approvable endpoint (a first for any species). What we admire most is the noble mission of giving dogs a longer leash on life, but a close second is the ingenious GTM. Building a biotech startup is excruciatingly difficult: founders face every risk at once (science, tech, regulatory, capital). And the grand prize looming beyond Loyal, human longevity, sits behind the tallest bar in the business: trials that would take decades and billions, against a regulatory regime that doesn’t recognize aging as a disease. So why not treat dogs first as your wedge? Trials run in dog years (~7x ours), vets provide a distribution channel for the drugs while confirmatory data comes in, and each stage of derisking builds a stronger case for the human version. Start with man’s best friend → end with man. Kudos to the Loyal team! (And… let’s chat.)
THE STREET ROUND ROBIN // while we love our VCs here at PA, we try not to overindex on early stages of equity financing, because as we’ve said more times than we can count, our renaissance in hard pursuits will require all comers from every corner of the capital stack. So we keep a lot of ‘sensors’ on Wall Street, and we got three pings this week that we’ve been thinking a lot about:
Apollo has a great note declaring the reindustrialization of America officially underway. The firm says that restoring the manufacturing-and-defense capital stock to its 2000s share of GDP will take ~$2T of fresh capital (while a return to ‘80s levels will require ~$6.5T). The private construction boom of late, while a great start, has mostly been about chips and compute — with the broader base barely moving. Upshot: time to get a move on everywhere else!
Morgan Stanley, like us and any of you knee-deep in the robotics supply chain, is actuator-pilled. In a Sep. 13 note titled Actuators: Robot Muscle Needs American Hustle, all-star Adam Jonas’s team, proxying actuators as motors + bearings + reducers, sizes the 2030 global TAM at $94B in its base case (en route to an eye-watering $3.8T by 2050).
A regulatory forcing function has arrived for the domestic market: the FCC now bars robots that don’t hit a 65% U.S. content threshold, rising to 75% by 2029.
Domestic capacity is nowhere close. Chinese actuator assemblies are 3x+ cheaper, and Chinese suppliers have grown their harmonic-reducer share from 41% to 88% since 2017.
We see this as the sell side underwriting the argument we made in #060: the merchant component base — the unglamorous middle of the stack — is where a decade of glorious opportunity and hard-tech margin will live.
JPMorgan’s commodities desk has thrown up its hands: “For the first time since the start of the Iran conflict, we don’t have a baseline view. We simply don’t know how to model the endgame.” The bank had assumed economic red lines ($100 oil, $5 gasoline, 5% Treasury yields) that would force an endgame. Each has since been crossed, and the conflict continues. This is a wider challenge our community (commodities desks, fuel buyers, metals procurement… anyone signing a multi-year offtake, really) faces today, with uncertainty at this macro level functioning like a tax on long-cycle projects.
“The Declining Economic Significance of US Manufacturing”, C/O Apollo - time to turn this ship around!!!
ATX HAPPENINGS // Austin Deep Tech Week is upon us and in full swing as we speak — check the schedule/grab tickets here, and if you’re in town, holler at Ryan — just reply to this email. He’s always game to grab coffee/tacos and talk actuator TAMs, launch cadence, why the Texas Triangle will be America’s Shenzhen, or whatever it is you’re obsessing over these days. And while you’ve got your calendar open: next Thursday, Oct 1, the Texas Venture Fest takes over downtown Austin from 5—8 PM. Between Overmatch Ventures, Array Labs’ TX office, and no shortage of PA regulars already calling the city home, this is about as close to a hometown crowd as we get outside our own events. Come meet the founders and funders building the next wave of Texas hard tech. Y'all won't want to miss it!
PUBLIC WORKS OF ART. Friend of PA Ryder Kimball spent last week pulling absolute stunners, then more bangers, out of the California Department of Water Resources' photo archive. Turns out the state water bureaucracy has its own Photography Unit on payroll, with staff shooters who have documented the State Water Project for decades, from Oroville's rebuilt spillway mid-release to the aqueduct snaking through the Central Valley to the snow surveys in the high Sierra. While we’re here…
Ryder leads content at Lowercarbon Capital, which backed Rainmaker’s $100M Series B this week (big shout-out to the rain-makers!)
He’s just published a dispatch from the cloud-seeders’ Alaska campaign, featuring drone-launching F-150s, moose hunters, and 19M gallons of new water in three hours, made from
thin airsupercooled clouds with less than a pound of silver iodide.Water abundance, or lack thereof, is a standing obsession around here: The End of Thirst Traps remains our definitive splash in the genre. Give it a read if you haven’t already.
Anyways, back to photography. There's a long American tradition here: Dorothea Lange was on a federal payroll when she shot the Depression, and the Bureau of Reclamation kept photographers at Hoover Dam, because somebody smart understood that if you want a country to believe in public works, you have to show people the works. Ryder had an idea we wanted to run by you — “give every state project a Photography Unit.” And honestly why stop there? Extend it to every big national build: the fabs, the shipyards, and yes, dare we say, even the datacenters.
Got thoughts? Write in and let us know what you think. Until then… go forth, do hard things, and we’ll see y’all back here in the inbox next week.
|