Large data
centers provide information processing, storage, and interconnection needed
to satisfy ever increasing demand for cloud
computing and artificial
intelligence. These facilities use massive amounts of electric power, water,
and real
estate measured in gigawatts, millions of gallons, and millions of square
feet respectively.
Not
long ago, data center ventures sought and received generous
financial inducements to locate in mostly rural locales in great need of
investment. Data centers seemed an attractive alternative to prisons, hazardous
material dump sites, and temporary housing for illegal aliens.
The Court
of Public Opinion on data centers has quickly pivoted from disinterest, or support
to “not in my backyard,” and even “not
on my planet you bastards.” Concerns
about data centers include environmental
harm, real or perceived upward pressure on retail
electricity, water,
and residential
housing prices, and likely
market domination of AI markets by a few “Big Tech” incumbents. Other
irritants include noise
pollution, and uncertainty
whether data centers support the local economy after the initial
construction phase.
Shifting
data centers from earth to space has become an
alternative touted as solving concerns without generating new ones. In sun
synchronous, low
Earth orbit, orbiting data centers (“ODCs”) may have
comparatively less environmental impact and lower
operating costs in two increasingly expensive categories: the gigawatts of
electricity continuously required to power components, and coolants, such as
water and glycol, needed to syphon away heat generated by constant operation.
Space
Commerce Exuberance
The
global consulting firm McKinsey & Company estimates the global
space economy will reach $1.8 trillion by 2035, up from their $630 billion
estimate for 2023. Low Earth Orbiting
(“LEO”) satellite constellations have the potential to provide reliable
infrastructure for widely
available and affordable broadband access, even in remote, rural, and
impoverished locales throughout the world.
Other developing
market opportunities include development of a vibrant space launch and
tourism industry, mineral extraction from asteroids, colonization of the Moon
and Mars, and an expanded array of services via commercial satellites.
Some of the world’s most innovative
and wealthy entrepreneurs believe space will serve as a profitable frontier
well worth the risk and expense. On the
issue of ODCs, Elon
Musk made a bold prediction: “I actually think that the cost of deploying
AI in space will drop below the cost of terrestrial AI much sooner than most
people expect. . . . I think it may be
only two or three years.”
Replacing terrestrial data centers
(“TDCs”) largely depends on declining launch costs and the ability to tap
“free” solar energy coupled with efficient radiational cooling of the heat
generated from continuous operations. To achieve a comparative advantage over
TDCs, ODC satellite constellations must collectively deliver services at or
below the per unit cost achieved terrestrially.
Are ODCs a Viable Extraterrestrial
Solution?
ODCs conceptually offer a more
environmentally friendly competitive
alternative to TDCs. On the other
hand, ODCs will complicate several key space management and governance issues. ODC
proponents may have underestimated or refrained from acknowledging significant
risks, increased by the massive increase in the number and close proximity of
operational LEO satellites. Potential launches
of LEO satellites numbering the millions raise the odds for collisions,
especially orbiting
in the narrow sun synchronous orbital plane.
Space
debris proliferation, without any significant mitigation efforts so far,
contributes to congestion,
higher collision risks, and the cascading effect of even more debris. At some point, the potential for calamity
reaches a tipping point where investors lose enthusiasm and confidence,
particularly when one or more collisions actually occur.
Additionally, the combination of
many more spacecraft launches, and deorbiting of decommissioned spacecraft,
will generate toxicity and pollution having uncertain impact in space and
Earth’s atmosphere. Consequently,
pro-market, regulatory forbearance and statutory interpretations may not
continue, particularly if the space commerce environment becomes more risky and
costly.
The Gravitational Pull from Unresolved Legal, Regulatory, and Space
Governance Issues
It remains
uncertain how many small LEO satellites will be needed to match the output
available from a current Walmart-sized terrestrial footprint and new hyperscale centers forecasted to have an even larger footprint, with power
requirements in the 1-5 Gigawatts range, equivalent to 1-5 nuclear power plants
and far more solar, wind, gas, oil, and hydroelectric facilities.
Assuming
space as a regulation free, or less regulated venue is highly unlikely. While
nation states may appear to lack jurisdiction over spacecraft that quickly
moves into and out of airspace, there are terrestrial anchors. No one questions the legality of national
regulatory authorities requiring applications to launch satellites, use radio
spectrum, and operate both domestically and internationally. Additional regulatory authorities can extend
to the nation where spacecraft launches occur, gateway and backhaul earth
stations operate, and intact space debris causes damage.
Vastly more
LEO spacecraft and “paper satellites,” not yet launched, if ever, will test
whether nations can effectively coordinate spectrum and orbital plane usage. As ODC constellations will operate within the
narrow parameters of sun synchronicity, they will have to convince regulators,
legislators, and investors that potentially millions of satellites, in
relatively close proximity, can avoid signal interference, collisions, and
orbital coordination conflicts.
ODC commercial prospects may trend down in
the degree of exuberance to cautious optimism based on more realistic assessment
of risks and rewards. A sizeable list of challenges includes increased
government oversight, coordination and other timetable delays, spacecraft
collision potential, whether AI will become profitable, and recognition that
operating in space does not eliminate all environmental and other fears,
uncertainties, and doubts.
