"Quantum computing is simply a probabilistic method... whereas quantum computing basically uses a probabilistic approach where it can be a variety of answers to get to an outcome" - Nate Mason [12:50]
"I know oftentimes people think of quantum computing or quantum as hardware and that's where a lot of money and energy has gone... but there's also the algorithm component and that's where we are." - Nate Mason [14:48]
Disclaimer: Orignal content owned by or sourced from third parties. It does not represent the views of 'Nuggets' platform or it's team. AI is used extensively across this platform including for summaries. Accuracy is not guaranteed, there can be mistakes. Any info or content on this platform is not a financial, legal, or investment advice. Do your own research. Refer for complete disclosures:- Terms of Use · Full Disclaimer
"It's not going to be just turning a switch on and now you're using quantum computers and quantum computing is going to solve everything... you're going to have a data center that has CPUs... GPUs and then QPUs." - Nate Mason [18:33]
"GPUs are not really being utilized to their maximum capacity... what that's causing is organizations to build more data centers... causing energy prices to soar." - Nate Mason [34:13]
"With quantum transfer learning we've been able to reduce some models by 99%... we were able to reduce it to 2,000 [parameters] which gave it a 99% reduction which now you can run this on edge." - Nate Mason [35:48]
"We're able to like have a 250x speed up... propagating the catalog... of over 3,000 satellites... in less than 1 second." - Nate Mason [40:51]
"If you want to order a pizza you're not going to use a quantum computer to do that... you're going to reserve the CPU or GPU for that." - Nate Mason [18:53]
Speakers & Credentials
Chris Gennady: Co-host of "The Next Big Thing" podcast. Focuses on technology and thematic topics at WisdomTree.
Elvira: Co-host of "The Next Big Thing" podcast, operating out of London. Collaborates with Chris on technology themes at WisdomTree.
Nate Mason: Head of Strategic Growth at BQP, a quantum algorithms company. Former intelligence professional with a 20-year career in the Air Force and NSA, focusing on electrical engineering and cryptography. Former venture capitalist focused on deep tech, space, and AI.
1. Executive Summary
The transition to quantum computing will not be an overnight switch but a hybrid evolution where CPUs, GPUs, and QPUs work in tandem within data centers to solve specific classes of problems.
The current bottleneck in computing is not just hardware, but outdated mathematics that limit GPU utilization, driving up energy costs and data center demand.
BQP is bridging the gap to the quantum era by developing "quantum-inspired" algorithms that run on classical hardware today, delivering massive performance gains for optimization and physics-based problems.
Quantum transfer learning is already demonstrating extreme efficiency gains in AI, reducing model sizes by 99% and enabling complex computations on edge devices (like satellites).
While the threat to existing cryptographic standards (like RSA 2048) by quantum computers is real, the immediate focus should be on practical problem-solving in sectors like aerospace logistics and space domain awareness, where quantum algorithms are already drastically outperforming classical methods.
A Non-Academic Entry: Nate Mason is not a quantum physicist; his journey began as an enlisted member of the Air Force post-9/11 [00:02:03].
NSA Experience: He spent much of his 20-year military career working alongside the National Security Agency (NSA), focusing on cryptography and signals analysis to protect deployed troops in Iraq and Afghanistan (specifically referencing battles in Fallujah) [00:07:25]. He noted the reality of the NSA is very different from Hollywood depictions like the movie Enemy of the State [00:06:16].
The VC Pivot: After obtaining his MBA at UC Berkeley (Haas), Mason transitioned into venture capital, investing in over 100 early-stage startups across deep tech, space, and AI [00:10:40].
The Appeal of Quantum: He was drawn to quantum computing because it functions as a foundational "back-end technology" applicable across multiple verticals (aerospace, defense, supply chain, biotech) rather than being siloed into a single industry [00:11:01].
Redefining Quantum Computing: Mason defines quantum simply as a probabilistic method, contrasting it with the deterministic (1 or 0) nature of classical computers [00:12:50].
The Math Bottleneck: While massive capital flows into quantum hardware, BQP focuses on the software layer. Mason argues that modern GPUs are underutilized because they rely on mathematics developed in the 1980s for CPUs [00:13:42].
Quantum-Inspired Computing: BQP develops algorithms using principles of quantum mechanics (superposition, entanglement) represented through tensors and matrices to dramatically improve the throughput of existing, classical hardware [00:14:05].
The Hardware Reality: Mason points out that a complex computational fluid dynamics (CFD) problem might require 19.2 million classical CPU cores to solve [00:16:00]. In contrast, the same problem could potentially be solved by a quantum computer utilizing just 30 logical qubits [00:16:16].
Shor's Algorithm: The hosts bring up the anxiety surrounding Shor's Algorithm (developed in 1994) and the impending threat to current cryptographic standards like RSA 2048 and elliptical curve cryptography [00:24:16].
Accelerating Timelines: Researchers continually publish papers revising down the number of qubits required to break encryption. Estimates have dropped from millions of qubits to 500,000, shrinking the perceived timeline to a critical vulnerability [00:25:17].
Pragmatic Defense: Mason remains level-headed, noting that transitioning from lab breakthroughs to industrial-scale attacks is incredibly difficult. Furthermore, institutions like NIST are already actively developing post-quantum encryption protocols [00:29:03].
Quantum as the Solution: He points out that the properties of quantum mechanics that pose a threat can also be used for defense. For instance, the principle of entanglement (where observing a particle alters it) could form the basis of unbreakable encryption methods [00:30:32].
The GPU Dilemma: Inefficient algorithms require organizations to build massive data centers and hoard GPUs, leading to skyrocketing energy prices and political friction [00:34:21].
Quantum Transfer Learning: BQP uses a technique called quantum transfer learning to drastically shrink AI models. Mason cites a computer vision model that was reduced from 14.5 million trainable parameters down to just 2,000—a 99% reduction in size [00:35:48].
Edge Computing Enablement: This extreme optimization allows complex AI models to run on edge devices, such as an Nvidia Jetson GPU, using significantly less power [00:36:07].
Real-World Use Case: Space Domain Awareness [00:37:42]
The Data Deficit: Tracking satellites is difficult when they pass over areas lacking sensor coverage (like the Pacific Ocean). Without constant data, organizations must predict orbital trajectories using "propagators" [00:39:21].
Unprecedented Speed: Working with the US Space Force's SDA TAP Lab, BQP developed a quantum-inspired orbital propagator that is 250 times faster than Orekit, which is the classical state-of-the-art system [00:39:50].
Scale and Accuracy: The new system propagated a catalog of over 3,000 satellites in less than 1 second [00:40:58]. It achieves accuracy under one kilometer in Low Earth Orbit (LEO) and between 15 to 25 meters in Geosynchronous Equatorial Orbit (GEO) [00:40:06].
Edge Deployment in Space: Because the resulting model is under one megabyte, it can theoretically be deployed directly onto a satellite in orbit via an Nvidia Jetson GPU, drastically reducing latency [00:41:20].
Engineering Optimization and the Future [00:51:34]
Breaking the Variable Ceiling: Classical optimization in aerospace design often limits engineers to solving for roughly 100 variables, forcing compromises across aerodynamics, propulsion, and mechanics [00:52:02]. BQP is pushing that limit to 10,000, with a goal of processing 1 million variables simultaneously using quantum approaches [00:53:13].
Logistics Breakthrough: In a fleet-level cargo optimization problem, BQP's algorithms reduced the computation time required to optimize the packing of hundreds of planes from four months down to just two hours [00:53:48].
The Hybrid Data Center: Mason envisions a future where computing stacks require zero structural changes to implement quantum. Developers will use a single line of Python or MATLAB code to route workloads to a backend solver that automatically toggles between CPUs, GPUs, and QPUs based on the specific problem being solved [00:56:47].
The Reference Vault
4. Data & Figures
Data Point
Value
Context
Timestamp
Computational Core Requirement
19.2 million
Number of CPU cores required to solve specific complex CFD challenges classically.
The Hybrid Compute Infrastructure: Mason's framework for the future of computing rejects the idea that quantum machines will replace classical ones. Instead, he models a tiered architecture where CPUs, GPUs, and QPUs coexist in the same data center. Workloads are dynamically routed to the most appropriate processor; classical chips handle standard operations (ordering pizza) while QPUs handle massive combinatorial optimization problems. [00:18:33]
Quantum Transfer Learning: A machine learning strategy where principles of quantum mechanics are applied to radically shrink the size of neural networks. By reducing the number of trainable parameters from millions to thousands (a 99% reduction), this framework enables complex AI models to operate efficiently on low-power edge devices rather than requiring massive data center compute. [00:35:20]
"Quantum Now" vs. "Quantum Next": A strategic framework for organizations approaching the quantum transition. "Quantum Now" involves implementing quantum-inspired algorithms on existing classical hardware today to train engineering teams and gain immediate optimization benefits. "Quantum Next" is the strategic preparation for the integration of physical quantum hardware. [00:15:07]
6. Anecdotes
The Reality of the NSA vs. Hollywood: Mason counters the public perception of the NSA (driven by movies like Enemy of the State). He notes his work was not domestic surveillance, but rather building mathematical tools and signals analysis programs to protect deployed soldiers during active combat in Fallujah, Iraq. [00:06:16]
Deploying AI to Orbit: Mason illustrates the power of quantum transfer learning by describing how BQP shrank an orbital prediction model to under 1MB. Because it is so small, they tested it on an Nvidia Jetson edge GPU, proving that highly complex orbital calculations can be physically sent to space and executed on the satellites themselves, removing the need for ground-based latency. [00:41:20]
Data Centers in Space: The hosts and Mason discuss the emerging reality of placing data centers in Very Low Earth Orbit (VLEO). Mason notes he saw pitches for this years ago. The logic is compelling: VLEO provides free cooling (the vacuum of space), free energy (solar), and minimal latency, offering a radical solution to the massive energy consumption of terrestrial AI data centers. [00:46:09]
7. References & Recommendations
Companies & Organizations
BQP: Nate Mason's current company, focused on quantum algorithms rather than hardware. Mentioned in the context of creating quantum-inspired math to maximize existing GPUs. [00:13:22]
National Security Agency (NSA): Mason's former employer, discussed in the context of his work in cryptography and signals analysis during the wars in Iraq and Afghanistan. [00:02:56]
IonQ: A publicly traded quantum hardware company known for trapped-ion modalities. Mentioned regarding their announcements on "World Quantum Day" causing market volatility. [00:21:19]
Nvidia: Mentioned frequently as a partner of BQP, a behemoth entering the quantum space via Ising models, and the manufacturer of the Jetson GPU used for edge computing tests. [00:32:23]
SpaceX: Used as the primary example of how commercial space launch costs have plummeted, enabling the rapid deployment of thousands of satellites. [00:44:02]
NIST (National Institute of Standards and Technology): Referenced as the organization actively developing post-quantum encryption protocols to defend against future quantum attacks. [00:29:03]
SDA TAP Lab: The US Space Force's "Space Domain Awareness Tools, Applications, and Process Lab," partnered with BQP to solve orbital propagation challenges. [00:38:14]
Anthropic (Claude): Mentioned briefly regarding AI companies holding back capabilities because their models are highly effective at finding zero-day cybersecurity vulnerabilities. [00:28:11]
Classiq: A quantum software company that partnered with BQP and Nvidia for hybrid benchmarking. [00:55:15]
Rigetti, IBM, Google: Mentioned as other major players developing specific, non-general-purpose quantum hardware. [00:27:32]
Geopolitical & Historical Events
9/11 (September 11 Attacks): The catalyst for Nate Mason joining the military in 2003. [00:02:03]
Battles of Fallujah: Referenced as the specific theater of conflict in Iraq where Mason's technical work was deployed to protect troops. [00:07:25]
Artemis Program: Mentioned alongside SpaceX as a major catalyst for making the space sector popular and driving commercial interest. [00:44:45]
Scientific & Technical Concepts
Shor's Algorithm: The 1994 algorithm that theoretical proves quantum computers can break standard cryptographic encryption (like RSA 2048). [00:24:16]
RSA 2048 & Elliptical Curve Cryptography: The current standards of internet encryption threatened by the advancement of quantum computing. [00:24:26]
Orekit: The current state-of-the-art open-source space dynamics library used as a baseline benchmark for BQP's orbital propagator speed tests. [00:39:50]
VLEO (Very Low Earth Orbit): Discussed as the ideal atmospheric layer for placing future data centers, offering a balance of space-cooling and low latency. [00:47:42]
Computational Fluid Dynamics (CFD) & Finite Element Analysis (FEA): The specific types of incredibly dense physics problems that require millions of CPU cores, making them prime targets for quantum replacement. [00:15:43]
Media & Pop Culture
Enemy of the State (Movie): A 1998 Will Smith film used as the quintessential example of the public's incorrect, paranoid perception of the NSA. [00:06:16]
World Quantum Day (April 14th): A recently established awareness day utilized by companies to make announcements and drum up interest in the sector. [00:20:54]
Dwarkesh Patel Podcast: Referenced regarding a specific episode where Elon Musk claimed data centers will move to space within three years. [00:46:09]
8. The Bottomline (by AI)
The true bottleneck in modern computing isn't just a lack of GPUs, but the reliance on outdated mathematics to run them—a problem that "quantum-inspired" algorithms are already solving today without the need for physical quantum hardware. While the media focuses on the existential cybersecurity threat of theoretical quantum machines, real-world operators are using these algorithms right now to radically shrink AI models (by 99%) and execute complex logistical and orbital computations in seconds rather than months. Organizations must shift their focus from fearing "Q-Day" to adopting hybrid algorithmic solutions immediately; those waiting for the physical hardware to mature will find themselves severely outpaced in engineering, logistics, and power efficiency.
"Brookfield's the largest infrastructure owner in the world... We drew a pipeline and we showed all the different components of the payments ecosystem on a pipeline and said it's like a pipe that moves any commodity except what it's moving…
Speed Increase (Space)
250x
The speedup of BQP's orbital propagator compared to the state-of-the-art (Orekit).