Latest / Elon Musk Podcast / SpaceX IPO Confirmed

Transcript

SpaceX IPO Confirmed · Elon Musk Podcast

• 0:00
  Elon Musk confirmed SpaceX will go public in 2026, reversing a
• 0:03
  position he held for more than a decade.
• 0:06
  The IPO could raise $30 billion in value of the company in 1.5
• 0:10
  trillion, making it the largest public offering in history,
• 0:14
  surpassing Saudi Aramco's $29 billion listing in 2019.
• 0:20
  Musk had long insisted SpaceX would remain private until
• 0:23
  rockets were flying regularly to Mars, arguing that public
• 0:27
  shareholders would prioritize profits over his colonization
• 0:31
  dreams. So why did he change his mind?
• 0:33
  The answer involves data centers in space.
• 0:36
  Musk plans to retrofit new generation Starling satellites
• 0:40
  into orbital computing platforms that could power the AI
• 0:44
  revolution, and he needs massive capital to make it happen.
• 0:47
  Today, we're going to cover the technical specs of these space
• 0:50
  data centers, the competitive race with Jeff Bezos and Google,
• 0:53
  and what this means for Musk's Mars timeline.
• 0:57
  Now, SpaceX confirmed the 2026 IPO after ours Technica reporter
• 1:01
  Eric Berger published an analysis of why the company
• 1:04
  would finally go public. Musk responded on X with four
• 1:07
  words. As usual, Eric is accurate.
• 1:10
  The company is currently conducting an insider share sale
• 1:13
  at an $800 billion valuation, exactly double what SpaceX
• 1:18
  commanded at its last capital raise in 2025.
• 1:22
  And the core driver behind this reversal is a new business line
• 1:26
  that would not exist when Musk first swore off public markets.
• 1:30
  SpaceX wants to build orbital data centers using its Starlink
• 1:33
  V3 satellites, which feature high speed laser links capable
• 1:37
  of 1 terabit per second throughput.
• 1:41
  Now, that number is 10 times the capacity of current Starlink V2
• 1:45
  satellites. The V3 satellites require
• 1:48
  Starship to launch because they are significantly larger, and
• 1:51
  SpaceX plans to deploy 60 of them per flight starting in
• 1:55
  2026. Unlike previous Starlink
• 1:58
  generations optimized purely for communications, the V3 units are
• 2:02
  being engineered as compute capable infrastructure with
• 2:05
  larger solar arrays, machine learning accelerators, and
• 2:09
  radiation shielding designed to extend operational life in the
• 2:12
  harsh space environment. Each satellite will process data
• 2:16
  before transmitting it to Earth, reducing bandwidth requirements
• 2:19
  and enabling real time AI inference at latencies below 10
• 2:24
  milliseconds. Now space offers several
• 2:27
  advantages over terrestrial data centers that make this bet
• 2:30
  rational. Solar panels in orbit capture
• 2:33
  roughly 8 times more energy than ground installations because
• 2:37
  there is no atmosphere, no weather, and no night time and
• 2:40
  the sun facing side. Cooling comes naturally through
• 2:44
  radiative heat dissipation into the vacuum of space, eliminating
• 2:48
  the massive water and electricity requirements that
• 2:50
  plague Earth based facilities. And the International Energy
• 2:53
  Agency reports that data centers consumed about 415 terawatt
• 2:58
  hours of electricity globally in 2024, roughly 1.5% of total
• 3:04
  power consumption. That figure is climbing rapidly
• 3:07
  as AI training runs demand more computational resources now.
• 3:12
  Goldman Sachs forecasts that global power demand from data
• 3:16
  centers will increase 50% by 2027, as much as 165% by the end
• 3:22
  of the decade. Orbital facilities avoid the
• 3:24
  land use conflicts and NIMBY opposition that have stalled
• 3:28
  terrestrial data center construction across the United
• 3:31
  States and Europe. Musk has described an even more
• 3:34
  aggressive long term vision. He wants to build satellite
• 3:37
  factories on the moon and use an electromagnetic rail gun to
• 3:41
  launch EI satellites at lunar escape velocity without rockets.
• 3:46
  In his estimation, his approach could eventually deliver 100
• 3:50
  terawatts of AI computing capacity per year into orbit.
• 3:54
  Figure sounds absurd until you remember that Musk has a track
• 3:57
  record of building things other people called impossible.
• 4:02
  Now SpaceX is not pursuing this opportunity in isolation.
• 4:06
  Jeff Bezos announced in October that Blue Origin has been
• 4:08
  developing orbital data center technology for more than a year.
• 4:12
  He predicted GW scale space data centers without 10 to 20 years
• 4:17
  and claimed they would eventually beat the cost of
• 4:20
  terrestrial facilities. Google launched Project
• 4:23
  Suncatcher in November of 2025, partnering with Planet Labs to
• 4:27
  put TPU chips on satellites launching in early 2027.
• 4:32
  And Google's research paper envisions 81 satellite clusters
• 4:35
  arranged in a 1 kilometer radius formation, communicating through
• 4:39
  free space optical links to distribute AI workloads across
• 4:43
  the constellation. The NVIDIA backs startup Star
• 4:46
  Cloud already has a satellite in orbit running an H-100 GPU,
• 4:50
  successfully training Google's Jemma A1 model in space.
• 4:55
  Now. Star Cloud CEO told CNBC the
• 4:58
  orbital data centers would have 10 times lower energy costs than
• 5:02
  terrestrial facilities. Company plans to build a 5 GW
• 5:06
  orbital data center with solar and cooling panels measuring
• 5:09
  roughly 4 kilometers in both width and height.
• 5:13
  Now China has entered the race with its three body computing
• 5:16
  constellation, which launched 12 satellites in May 2025 through a
• 5:20
  partnership between Zhengjiang Lab and Alibaba Group.
• 5:25
  The financial picture for SpaceX makes the IPO timing very
• 5:29
  logical right now. Company generated roughly $15
• 5:32
  billion in revenue in 2025, is projected to hit about 24
• 5:36
  billion in 2026, putting it roughly on par with NASA's
• 5:40
  annual budget. And Starlink alone has grown to
• 5:43
  over 9 million subscribers from 4 million just 15 months
• 5:47
  earlier. Musk stated publicly the less
• 5:50
  than 5% of SpaceX revenue now comes from NASA, meaning the
• 5:53
  future of the company depends almost entirely on Starlink.
• 5:57
  SpaceX launched more than 160 Falcon 9 rockets in 2025, which
• 6:01
  is more than the rest of the world combined, and put over
• 6:04
  3200 Sterling satellites into orbit during the calendar year.
• 6:08
  A single Falcon booster, designated B1O67 has now
• 6:12
  launched and landed 32 times, with turn around windows as
• 6:16
  short as three weeks. The company hit successful 500
• 6:19
  landings in October 2025, more than every other company in
• 6:23
  nation state combined. Blue Origin, in second place,
• 6:27
  has landed exactly 1. Now Mars, it remains on the
• 6:33
  forefront. It's the elephant of the room
• 6:36
  for long time SpaceX fans. Musk founded the company
• 6:39
  specifically to make humanity multiplanetary, and he spent
• 6:42
  years arguing the public shareholders would never
• 6:45
  tolerate that expense. Now he said it would be
• 6:48
  unprofitable work of colonizing another planet.
• 6:51
  In 2023, he told Twitter followers that SpaceX would only
• 6:54
  go public when Mars colonial transport, Now SpaceX President
• 6:59
  Gwyn Shotwell reinforced his position in 2018, saying the
• 7:02
  company would remain private until rockets were flying
• 7:05
  regularly to the Red Planet. That's not happened yet.
• 7:08
  They haven't even gotten a Starship to orbit yet.
• 7:12
  There's no in orbit refuelling, and they haven't landed on
• 7:15
  another celestial body, let alone on Mars.
• 7:19
  Starship has not completed a successful orbital mission after
• 7:22
  several explosive tests, and unaccrued flights to Mars remain
• 7:25
  at least a year away by the most optimistic projections.
• 7:30
  But that's probably not going to happen as well.
• 7:31
  Elana stated on the record that Mars is the distraction now.
• 7:36
  Justice Palmer, CEO of Fortuna Investments and venture capital
• 7:40
  firm invested in SpaceX, said he expected an IPO to happen after
• 7:45
  Starship reaches Mars because it would remove a massive risk
• 7:49
  factor. A failed Mars attempt as a
• 7:52
  private company causes no risk, no stock fluctuation, but the
• 7:56
  same failure as a public company could hammer the share price.
• 8:01
  Now the technical challenges are substantial, but not
• 8:05
  insurmountable. Radiation in space can damage
• 8:07
  computer chips, causing bit flips and burnout that corrupt
• 8:11
  AI calculations. Space rated processors currently
• 8:15
  lag three to four orders of magnitude behind commercial
• 8:18
  hardware and raw performance, but Google estimates that space
• 8:21
  based AI computing will only become economically viable by
• 8:25
  2035 if launch cost dropped to $200 per kilogram from the
• 8:31
  current 1400 to $2900 range. Starship is the key to hitting
• 8:36
  that target, which puts SpaceX in a stronger position than any
• 8:41
  other competitor.