a16z Podcast | Quantum Computing, Now and Next

Moore's Law and Classical Computing Limits:

• Transistor sizes have shrunk to approximately 10–20 nanometers, roughly the width of 100 atoms, representing a fundamental physical limit for classical silicon-based scaling.
• Power density and heat extraction have become critical bottlenecks as transistor density increases, leading to diminishing returns in energy efficiency.
• Amdahl's Law limits parallel computing efficiency; adding processors yields diminishing returns because a portion of computations must remain serial.
• The capital cost to build a fabrication facility (fab) for 10–20 nanometer nodes now exceeds $10 billion, restricting the ability to innovate to only a few global entities.
• Industry focus has shifted from brute-force hardware scaling to specialized architectures, including GPUs, FPGAs (e.g., Microsoft Azure), and neuromorphic chips, to accelerate machine learning.

Quantum Computing Fundamentals:

• Quantum computers differ from classical systems by encoding information in continuous quantum variables rather than discrete binary digits (0s and 1s).
• Computational capacity scales exponentially: adding a single qubit doubles the number of accessible continuous variables, unlike the linear scale of classical transistors.
• The technology relies on quantum coherence, the ability to maintain quantum states long enough to perform reliable calculations, a challenge first addressed in the early 2000s.
• Current hardware challenges involve packing sufficient qubits onto a single chip while maintaining low error rates for quantum Boolean operations.

Strategic Applications and Value Proposition:

• Computational Chemistry: Simulating molecular structures and materials properties using the Schrödinger equation, solving problems infeasible for classical supercomputers due to combinatorial complexity.
• Optimization and Machine Learning: Running hybrid quantum-classical algorithms where quantum processors optimize loops within classical hardware to solve complex optimization problems.
• Cryptography: Shor's algorithm could theoretically factor large numbers to break RSA encryption, though practical application is estimated 20–30 years away.
• Forward-looking vision suggests quantum computing will act as a co-processor (P-P-U) in a heterogeneous cloud environment, integrating seamlessly with existing classical infrastructure.

Industry Landscape and Competitive Dynamics:

• The field has evolved from a niche physics research pursuit to a global race involving major players (IBM, Google, Microsoft) and significant academic centers (ETH Zurich, TU Delft, institutions in Australia and China).
• Rigetti Computing positions itself as a full-stack startup, designing superconducting qubit integrated circuits, manufacturing chips, and developing software stacks for cloud integration.
• The CEO argues that startups hold a competitive advantage over incumbents by curating specialized "DNA" and avoiding the cultural inertia of large organizations, analogous to Tesla's early position against General Motors.
• The industry is currently in a "Cambrian explosion" where smaller entities can invest roughly $10 million in infrastructure to build chips rivaling parts of supercomputers built on $4 billion fabs.

Ecosystem and Future Outlook:

• The number of quantum software engineers is projected to grow from near zero to a meaningful cohort over the next five years, driving a "flywheel" of application discovery.
• The primary immediate bottleneck is no longer hardware but the development of sophisticated algorithms and software tools to leverage existing quantum resources.
• The technology is expected to reshape healthcare (drug discovery), energy generation, and agriculture through simulation-driven design, rather than just cryptography.
• A new "Silicon Valley" ecosystem is anticipated to emerge specifically around quantum computing technologies, distinct from the classical semiconductor industry.