Aggregator

Goldman Sits Down With Anduril As 'War Unicorns' Reshape Defense Tech

Zero Rss
2 weeks 2 days ago
Goldman Sits Down With Anduril As 'War Unicorns' Reshape Defense Tech

Palmer Luckey’s defense startup, Anduril, is emerging as the Department of War’s answer to the urgent need for affordable, scalable advanced weaponry produced at lightning speed, rather than through the slow, over-budget procurement cycles that have long defined the legacy primes.

The twin conflicts raging across Eurasia and the Middle East, from the Russia-Ukraine war to the U.S.-Iran war, have forever altered modern warfare, with drones, seaborne drones, ground robots, and AI kill chains now reshaping the battlefield.

The quick rise of Anduril, something we call a “war unicorn,” has attracted the attention of Goldman analysts, who recently felt compelled to sit down with Anduril executives to better understand the story and how it will play a major role in the next phase of rebuilding America’s defense-industrial base.

Analyst Noah Poponak recently hosted Anduril co-founder and CEO Brian Schimpf and head of investor relations Allison Lazarus in New York to gain more color on how the defense company is solving the defense industry’s biggest bottleneck, speed.

Oculus headset creator Palmer Luckey, who founded the company in 2017, has focused on building lower-cost, scalable systems in categories such as drones, counter-UAS, and missiles, positioning itself against a legacy defense-industrial base that includes Lockheed Martin, Boeing, and many others.

Here are Poponak's top takeaways after speaking with Anduril executives:

What is Anduril solving for? The U.S. defense industrial base is currently geared towards producing low numbers of expensive, bespoke assets. While these assets are very capable - they have extremely high specifications and performance requirements - they have historically been used in limited quantities, often utilize sole-source specialty materials and components, and have complex manufacturing processes, all of which makes scaled production ramp-ups difficult. More recently, the rate at which these assets have been used in modern conflicts relative to their respective manufacturing footprints and stockpile levels has proven to be high. This has prompted the United States to 1) increase the production rate of these assets, but 2) develop and procure lower-cost alternatives that can be mass-manufactured. This is where Anduril steps in. The company designs its products with affordability and scalability in mind ("affordable mass").

In the long-run, Anduril believes its capability set can be a differentiator, but right now it is prioritizing scalability and speed by focusing on these principles:

1. Vertical integration. When it makes sense to do so, Anduril brings component and module production in-house, decreasing its reliance on external vendors to meet production targets. When the company does rely on external vendors (e.g. Anduril owns the design for a component, but does not produce it in-house), it seeks strategic relationships with commoditized component manufacturers, aiming to avoid the supply chain risks sometimes posed by sole-source suppliers (although this is hard to avoid with modules). Importantly, Anduril is not looking to completely vertically integrate. When the industry has many suppliers of a common item (such as small turbo jets or bolts), Anduril will source that item externally.

2. Component and process commonality across product lines. By utilizing the same components (bearings, bushings, washers, electronic components, materials) across product lines, Anduril simplifies its supply chain and can flex a common inventory pool across products. Having similar production processes allows for an easier manufacturing switch between products.

3. Product simplification and production automation. By designing simpler products, Anduril creates more opportunities for automation, lowers cost, and enables higher production output. A simpler product often times comes with a lower capability set, but there are increasing use cases for these products, as best-in-class assets are not always needed to achieve mission outcomes.

4. Flexible manufacturing facilities. By designing production lines and facilities to rapidly switch between different products, Anduril is able to more quickly meet demand when and where it occurs. This also lowers the risk of idle, dedicated manufacturing space for products not currently in a procurement cycle.

Internal investments create product adoption and margin upside. In our meeting, Anduril stated a near 25% total company operating margin over time is possible. The company derives a high percentage of its revenue from fixed-price work (we estimate that to be between 70-80%), and invests substantially in internal R&D, which, per the company, could allow it to command much higher margins than traditional defense primes over time. Investing ahead of customer demand does entail risk, but the company believes it is advantageous to compete with actual in-production assets versus hypothetical assets or designs when industry RFPs are released, combined with an ability to deliver to schedule. The DoW is increasingly requesting that industry participants invest their own capital into R&D and production, as it fosters faster product iteration and more competition.

DoW acquisition reform - showing early signs of progress. While acquisition processes at the DoW are still complex and can be difficult to navigate, Anduril stated it is seeing improvement at the customer, noting an increased willingness to try new testing, prototyping, and procurement strategies across the services. At an industry-wide level, the DoW is implementing acquisition reform (see our notes here and here), negotiating multi-year framework agreements, prioritizing fixed-price contract terms, investing in companies via equity stakes, and establishing more open test and procurement strategies like the Drone Dominance Program.

What products in the portfolio are seeing momentum?

  • Anduril's Fury (YFQ-44A) continues to progress through the USAF's Collaborative Combat Aircraft (CCA) program as it performs aerial testing and enters the production phase at Arsenal-1.
  • The company's Dive-XL product is being manufactured for the Royal Australian Navy's Ghost Shark XL-AUV program.
  • Anduril recently entered into a framework agreement to produce thousands of Barracuda low-cost cruise missiles in the coming years.
  • Anduril's CUAS products continue to gain momentum domestically (USMC) and internationally (to protect critical infrastructure abroad).
  • Lattice is critically integrated into Anduril's hardware, but is also used for enterprise-wide / system-wide applications in the military (U.S. Army, U.S. Space Surveillance Network).

Anduril portfolio

Anduril notable awards and programs of record

Anduril fundraising rounds and implied valuation

Anduril M&A timeline

Professional subscribers can read the full Anduril note here at our new Marketdesk.ai portal. 

Tyler Durden Wed, 06/03/2026 - 11:35
Tyler Durden

Quantum Computing: Hype Or The Real Deal?

Zero Rss
2 weeks 2 days ago
Quantum Computing: Hype Or The Real Deal?

Authored by Michael Lebowitz via RealInvestmentAdvice.com,

The word “quantum” is defined as “an amount,” with its Latin root meaning “how much.” In physics, a quantum is the smallest discrete unit of any physical property. The adjective form, as in “quantum leap,” describes a sudden, significant, and fundamental change. It is also a perfect adjective for coining the next potential technological innovation—quantum computing.

Quantum computing is being hailed as the next technology to revolutionize computing, following in AI’s footsteps. The headlines below, though confusing for non-quantum experts, are quite impressive:

  • Google Unveils ‘Mind-Boggling’ Willow Chip, Solving 10-Septillion-Year Problem in Minutes

  • World’s First 10,000-Qubit Processor Achieves 100x Leap in Qubit Count

  • JUPITER Supercomputer Breaks World Record with 50-Qubit Quantum Simulation

  • Caltech Physicists Build Massive 6,100-Qubit Neutral Atom Array

  • A Photon Was Teleported Across 270 Meters in Stunning Quantum Breakthrough

But promising headlines loaded with industry jargon have a long history of appearing well ahead of reality. Accordingly, let’s better understand what quantum computing is, why it matters, and whether the hype is justified and worth investing in.

Classic Computing & Moore’s Law

To understand quantum computing, we need to first understand what bits are.

Every classical computer, from your smartphone and the laptop on your desk to El Capitan, the world’s most powerful supercomputer, operates using bits.

A bit is the fundamental unit of information: it is either a 0 or a 1. Every email you send, video you stream, and game you play is the result of billions of 0s and 1s switching on and off.  Over the last 50+ years, innovations have continually optimized the performance of these binary operations. However, continuing down this path becomes increasingly harder.

Moore’s Law describes computing innovation, while the law’s limits best describe the problem. Moore’s Law states:

The number of transistors on a microchip doubles roughly every two years, while the cost of computers is cut in half.

As the graph below shows, Moore’s law has resulted in exponential growth in computing power at increasingly lower costs. Note that the y-axis is on a log scale, with each tick representing a 100 times improvement in computing power per dollar.

The problem with Moore’s law is that it is limited by physical barriers. Specifically, the ability of chip manufacturers to continually shrink transistors. While there have been many innovations that extend the lifespan of Moore’s Law, the limits dictated by physics are becoming harder and more expensive to overcome.

Quantum Bits

Quantum computing takes a fundamentally different approach than the binary bits used by classical computers. Instead of bits, it uses quantum bits, or qubits.

Unlike bits, which exist only as a 0 or a 1, a qubit can be a 0 and 1, but it can also exist in three additional properties as follows:

  • Superposition: A qubit can exist as a 0 and a 1 simultaneously, allowing a quantum computer to explore multiple solutions at the same time.

  • Entanglement: Two qubits can become linked such that the state of one instantly determines the state of the other, regardless of the distance between them.

  • Interference: Quantum algorithms exploit wave-like behavior to amplify correct answers and cancel out incorrect ones, steering the computation toward the right result.

Confusing? Absolutely, but the critical takeaway is relatively simple.  Quantum computing doesn’t process information sequentially. Instead, it explores many possible solutions simultaneously.

To help better appreciate the difference, let’s consider the task of solving a maze. A classical computer will try one path and, if it fails, try a second, third, fourth, and so on, until it stumbles upon the correct one. A quantum computer explores all possible paths at the same time, thus finding the correct path much more quickly.

For certain applications, such as cryptography, drug discovery, materials simulation, and financial optimization, where the number of potential “paths” is astronomical, quantum computing can significantly reduce computation time. To wit, consider the stunning quote below, courtesy of the New York Times:

Excitement Vs. Reality

Now we must quell the excitement and explain why patience is warranted.

Traditional bits, 0s and 1s, are robust. For instance, they operate efficiently regardless of room temperature or vibrations, like a Wi-Fi signal.

Conversely, qubits are fragile and need to be isolated from the outside world. Any contact with its environment, even something as subtle as a slight temperature change or a vibration, causes it to behave like a classical bit and lose its quantum characteristics.

Accordingly, quantum processors are cooled to temperatures near absolute zero, roughly -460 degrees Fahrenheit. For context, that is about 100 degrees Fahrenheit colder than Pluto’s typical temperature range.

To overcome these obstacles, they are housed in elaborate dilution refrigerators and shielded from interference. Even with these expensive controlled conditions, qubits can make errors at rates that would be unacceptable in classical computing.

Quantum Reliability

The quantum industry measures progress toward reliability using the ratio of logical qubits to raw physical qubits.  A logical qubit is one that has been error-corrected to behave reliably, as opposed to a raw physical qubit, which is prone to mistakes. It currently takes somewhere between 1,000 and 10,000 physical qubits to produce a single reliable logical qubit. To put that in perspective, useful quantum computing requires thousands of logical qubits. Therefore, the total physical count of qubits needs to be in the tens of millions.

Increasing the number of logical qubits is a massive task that engineers are working to overcome.  Accordingly, many researchers think that fault-tolerant machines capable of solving real-world problems are still many years away.

The challenge in building qubits lies in two competing demands. They need to be isolated from the environment to maintain superposition and entanglement. But at the same time, practical enough to build, control, and scale into the millions needed for a useful computer.

No one has figured out the best way to build a qubit; accordingly, scientists are at various stages of developing numerous qubit types.

The graphic below, courtesy of Aliro, summarizes three approaches.  

The answer may be that a hybrid approach or something not yet in development overcomes these challenges.

Investing In Quantum

Despite the potentially long timeline and technological hurdles, the industry is making substantial progress. Accordingly, the investment possibilities are slowly coming into focus.  To help you get started with your research, we provide a snapshot of the publicly traded quantum computing companies.

Keep in mind that IBM, Alphabet, and Microsoft are large companies with numerous streams of substantial revenue. While those cash flows help fund quantum R&D, the ultimate impact of quantum computing on their bottom lines will be diluted by other business lines. IonQ, D-Wave, and Rigetti are quantum-centric companies. Any significant breakthroughs could be extremely valuable to shareholders. However, while they have some revenue to fund R&D, they will be much more dependent on debt and dilutive equity offerings. 

IBM (IBM)

IBM is arguably the most experienced builder in the field. Its Quantum roadmap continues to evolve. The 2026 roadmap below, courtesy of IBM, shows the company is targeting fault-tolerant systems with thousands of logical qubits by 2033. IBM is not solely a quantum computing company. Cash flows from its mainframe and hybrid cloud businesses help fund quantum research and development.

IBM has actively researched quantum computing since the 1970s and launched the IBM Quantum Platform, the first accessible quantum computer on the cloud in 2016. Their long-term commitment and prior successes provide them with a knowledge and infrastructure advantage over their competitors.

Hot off the presses: IBM and the US Commerce Department announced the US’s first purpose-built quantum foundry, supported by a proposed $1 billion chips grant. The funds are part of a $2 billion total going to nine quantum computing companies.

Google (GOOG)

Google achieved one of the field’s most important milestones in December 2024 when its Willow chip surpassed a key threshold. The chip demonstrated what the company called below-threshold quantum error suppression, meaning that adding more qubits reduced error rates rather than compounding them as was the case. In October 2025, Google announced “verifiable quantum advantage,” claiming its Willow chip completed a specific algorithm roughly 13,000 times faster than classical supercomputers. Google’s deep integration with DeepMind and its classical AI capabilities gives it a unique hybrid research platform.

Microsoft (MSFT)

Microsoft has pursued a different technical path, betting on topological qubits. While these qubits are much more stable than those in other approaches, they are incredibly difficult to create and verify in a laboratory setting. Microsoft released its first topological qubit chip in early 2025. Azure Quantum, its cloud platform, has become an important interface layer connecting users to multiple quantum hardware providers, including IonQ and Quantinuum.

IonQ (IONQ)

IonQ is the most prominent pure-play quantum company in the public markets. Rather than producing superconducting qubits, IBM and Alphabet’s approach, IonQ uses trapped-ion technology, in which individual ytterbium ions serve as qubits. The approach runs more slowly but is generally more accurate. In 2025, IonQ became the first quantum company to exceed $100 million in GAAP revenue. It has also been aggressively acquisitive, purchasing Oxford Ionics for approximately $1.1 billion and adding quantum sensing and networking capabilities through several smaller deals.

D-Wave (QBTS)

D-Wave takes a narrow but commercially pragmatic approach. Rather than pursuing universal quantum computing, D-Wave specializes in quantum annealing, a technique to optimize problems in logistics, supply chains, and scheduling. Its products may appeal to a relatively small number of clients, but it is the one generating the most near-term commercial traction. D-Wave’s Advantage systems are in active production use with real enterprise customers today, not just in a lab.

Rigetti Computing (RGTI)

Rigetti is a smaller pure-play competitor focused on superconducting systems. Its Ankaa-3 processor has shown improved interconnect performance. It has faced financial pressure, but it retains a dedicated engineering team and a growing cloud customer base.

Summary

The quantum computing story is real. The question is not whether we will see quantum computers widely used, but when. The underlying physics is sound, the engineering progress is documentable, and the potential applications are significant.

The investment case requires patience and a realistic view of the potentially long timeframe. By most industry estimates, fault-tolerant quantum computing capable of broadly outperforming classical computers on a commercially viable basis is, at least, a decade away. The companies best financially positioned to survive that long a wait are those with diversified revenue streams (IBM, Alphabet, and Microsoft) or near-term niche players like D-Wave.

Given the long development time horizon, the associated financial burden, and the uncertainty about which qubit types and companies will be the winners, we recommend a portfolio approach. A diversified approach ensures greater stability and improves the odds of success. Last, and maybe most important, patience is required, as quantum computing companies, especially the smaller ones, are likely to have fits and starts as they progress.

Views expressed in this article are opinions of the author and do not necessarily reflect the views of ZeroHedge.

Tyler Durden Wed, 06/03/2026 - 11:15
Tyler Durden

zero rss

Copyright (c) 2026 FYCKL Project