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

Episode Overview

• This episode features a panel from the All-In Liquidity Summit 2026, featuring Brad Gerstner (Altimeter Capital), Will Marshall (Planet Labs), and Andrew Feldman (Cerebras Systems) discussing the challenges and opportunities of taking deep-tech companies public.
• It frames the contrasting paths to the public markets, comparing Cerebras' traditional IPO to Planet Labs' SPAC merger, and explores the operational realities post-public listing.
• The discussion shifts into forward-looking technological paradigms, analyzing the feasibility of space-based data centers and the architectural shifts required in AI silicon.
• This content is highly relevant to investors, founders, and tech enthusiasts interested in hardware-scaling bottlenecks, space technology, and public market dynamics for venture-backed companies.

Key Concepts

• The Credibility Premium of Public Markets: Going public is more than just a liquidity event; for deep-tech companies selling to governments and massive enterprises (like agricultural giants or defense agencies), being a public company acts as a vital signal of longevity and financial solvency.
• Space as the Next Frontier for High-Power Compute: Space-based data centers are becoming economically viable due to two converging trends: the 10x reduction in launch costs (driven by SpaceX and reusable rockets) and the rapid miniaturization of satellites. Space offers a massive advantage in power generation because satellites in sun-synchronous orbits can access uninterrupted 24/7 solar energy without the need for expensive batteries or terrestrial cooling infrastructure.
• The AI Memory Wall (Memory-to-Compute Bottleneck): Traditional GPU architectures are bottlenecked by the physical distance data must travel between memory chips and processing cores. To solve this, new hardware architectures (like Cerebras' wafer-scale engines) integrate memory directly next to the compute on a single massive silicon wafer, vastly accelerating processing speeds for large language models.
• The Shift Back to Early Public Offerings: The venture capital paradigm of "staying private longer" is swinging back. While mega-companies like SpaceX and OpenAI capture massive private valuations, public markets allow a broader base of investors to capture the majority of a company's hyper-growth phase (the "5x to 10x" valuation curve) rather than having all upside captured solely by private investors.

Quotes

• At 3:21 - "You do all this work... and the next morning you've sold no more stuff, your engineering projects have made no progress since the day you weren't public, and you go back to work." - Andrew Feldman, explaining the grounding reality that going public doesn't automatically solve core operational and product development challenges.
• At 5:20 - "Everything was really hard until it got really easy. Like, nine and a half years of really hard and then twelve months of really easy..." - Brad Gerstner, describing the grueling, decade-long journey deep-tech startups face before suddenly reaching a market tipping point where everyone wants to invest.
• At 8:57 - "The maturing event gives you more credibility to various customers... they want to know you're going to be around." - Will Marshall, highlighting how being a public company is a crucial trust signal for risk-averse enterprise and government buyers.
• At 13:24 - "We are rebuilding the data processing infrastructure that has existed on the Earth, in the sky." - Chamath Palihapitiya, framing the massive macro-trend of shifting data capture and processing from ground-based systems directly into space-based orbits.
• At 13:25 - "The hard part here, the hard part is moving data from memory to compute. This is the fundamental problem in AI." - Andrew Feldman, explaining the physics-based bottleneck of modern silicon design and why architectural innovation is required to outpace traditional GPUs.

Takeaways

• When evaluating deep-tech investments, analyze the target customer base; if they sell to enterprise or sovereign entities, prioritize companies that pursue early public listings as a strategy to unlock commercial credibility.
• Look beyond raw transistor density and focus on chip architecture—specifically memory bandwidth and proximity—when assessing emerging AI silicon startups aiming to compete with dominant market leaders.
• Track launch cost trends (dollars per kilogram) as a leading indicator for the commercial viability of off-planet infrastructure, aiming to enter the space-compute market as launch costs dip below the $300/kg threshold.