The IPO Comeback: Why Tech Giants Are Finally Going Public | All-In Liquidity IPO Panel

IPO Market Outlook for 2026:

• Andrew Feldman (CEO of Cerebras) predicts 2026 could be an "all-time record" year for Initial Public Offerings (IPOs).
• Cerebras Systems is highlighted as the potential "AI IPO of the year."
• Will Marshall (CEO of Planet Labs) notes that the market is shifting back toward public listings after a decade of "stay private forever" strategies advocated by figures like Marc Andreessen.

Cerebras Systems IPO Performance & Strategy:

• The IPO pricing was set at $185 per share, with the range covered twice.
• The stock opened at $320 per share; current valuation sits near $230 per share with a market cap of $50–60 billion.
• The path to IPO was characterized by extreme difficulty for 9.5 years, followed by a 12-month period where "everything got really easy."
• Key investors include Capricorn, Peter Thiel's Founders Fund, and Yuri Milner's DST, with the UAE government also being a significant investor.
• Cerebras utilized an innovative "dribble lockup" agreement with banks, allowing shares to be released gradually over six months based on performance hurdles.
• Feldman advises investors to hold shares post-lockup, citing data that the majority of value creation for venture-backed companies occurs after the IPO.

Planet Labs IPO Performance & Strategic Pivot:

• Planet Labs stock has experienced a 10x increase over the last 12 months, rising from $5 to $50 per share.
• The company operates the largest Earth imaging fleet with approximately 200 satellites capable of imaging the entire Earth daily.
• Approximately 60% of Planet Labs' revenue is currently derived from security and defense customers.
• Marshall states that being a public company provides essential credibility to government and enterprise clients, reassuring them of long-term operational continuity.
• The company's valuation model relies on the convergence of space-based data and AI, creating "planetary intelligence" rather than just a data source.

The Convergence of AI and Space:

• Launch costs have decreased 10x over the last decade; current costs are roughly $1,000/kg, with a target of $200–$300/kg to make space-based data centers economically viable.
• Marshall projects that launch costs will hit the $200–$300/kg threshold within two to three years due to Starship technology.
• Space-based data centers could be powered by solar panels in dawn-dusk orbits, capturing 5x more energy per panel than terrestrial setups without needing batteries.
• Marshall estimates a $75–100 billion market for Earth observation data alone, excluding broader AI applications.
• Google and Planet Labs are partnering to launch TPUs into space, while Nvidia GPUs have already been deployed on Planet satellites for testing.
• Marshall predicts that within 10 years, a majority of computing infrastructure could be situated in space, creating a market valued in the trillions.

Cerebras Architecture & Competitive Landscape:

• Cerebras distinguishes itself by using a chip the size of a dinner plate (Wafer-Scale Engine), whereas most competitors use postage-stamp-sized chips.
• The architecture places memory directly adjacent to compute, solving the "data movement" bottleneck that limits traditional GPUs.
• Cerebras claims to be 15–18 times faster than NVIDIA GPUs for specific AI workloads, enabling real-time AI responses rather than delayed processing.
• Feldman argues that AI created a new class of problems (image and language processing) that required a fundamental rethinking of silicon architecture, rendering existing GPU designs insufficient for next-gen efficiency.
• Cerebras has achieved 20x performance gains by avoiding the "low-hanging fruit" optimizations that GPUs have already exhausted.

Investor Sentiment & Market Dynamics:

• Planet Labs serves as a case study where 10x returns were generated after the company went public, supporting the thesis that early public listings benefit public market investors.
• Will Marshall warns that for SpaceX to justify a future public listing at a quadrillion-dollar valuation, the market must anticipate a specific "takeoff" in applications.
• Panelists agree that public scrutiny and the discipline of public markets sharpen innovation and focus, countering the notion that private status is always superior for tech giants.
• Brad Gerstner (Altimeter Capital) views the AI/Space trend as the "biggest wave in the history of technology," predicting significant benefits for the US economy.

Future Outlook & Narrative Shifts:

• The industry is moving from viewing space as a niche data provider to a critical component of global infrastructure for agriculture, energy, and security.
• Marshall introduces the concept of "Large Earth Models," which would train AI on real-world Earth data to solve practical problems like flood prediction, agricultural planning, and conflict prevention.
• The "blind" nature of current text-based LLMs is being addressed by integrating live satellite data to provide real-world context to AI models.
• A significant shift in capital allocation is expected as venture capitalists and LPs recognize the potential for massive returns in public space and AI companies.