Atom Computing

Neutral Atom Founded 2018 Berkeley, CA, USA

Overview

Neutral atom quantum computers using optical tweezers. Focus on scaling to 1,000+ qubits with reconfigurable connectivity and long coherence times.

Current System: 1225 qubits
Funding: Private, raised ~$200M (Series C)
Website: https://atom-computing.com

Key Milestones

  • 2018: Atom Computing founded by Ben Bloom and Jonathan King
  • 2021: 100-qubit neutral atom system demonstrated
  • 2022: Partnership with DARPA for quantum networking
  • 2023: 1,180-qubit system (largest gate-based quantum computer)
  • 2024: 1,225-qubit Phoenix processor with Rydberg gates

Technology: Neutral Atoms

Atom Computing uses strontium-87 atoms trapped in optical tweezers (focused laser beams). Atoms are cooled to near absolute zero, arranged in 2D/3D arrays, and manipulated with lasers.

Advantages:

  • Long coherence times (~40 milliseconds, 1000x longer than superconducting)
  • Reconfigurable connectivity (atoms can be moved mid-computation)
  • Parallel gate operations (Rydberg blockade enables multi-qubit gates)
  • Scaling path (optical tweezers can trap thousands of atoms)

Challenges:

  • Slower gates (~1 μs vs. <100 ns for superconducting)
  • Complex optics (thousands of laser beams, precise control)
  • Loading time (seconds to prepare atom array)

Phoenix Processor

1,225 qubits arranged in reconfigurable 2D lattices. Uses Rydberg gates (atoms in highly excited states interact strongly) for entanglement.

Key metric: Gate fidelities ~99.7% (approaching error correction thresholds).

Competitive Position

vs. QuEra:
Both neutral atom companies. Atom Computing uses strontium; QuEra uses rubidium. Similar architectures, competing on qubit count and gate fidelity.

vs. Superconducting (IBM, Google):
Longer coherence, reconfigurable connectivity, but slower gates. Neutral atoms may win for error correction; superconducting for near-term NISQ algorithms.

Applications

  • Quantum simulation (many-body physics, condensed matter)
  • Optimization (QAOA with long coherence enables deep circuits)
  • Error correction experiments (long coherence reduces overhead)

Partnerships: DARPA, US Air Force Research Lab, Lawrence Livermore National Lab.

Roadmap

Atom Computing targets 10,000+ qubit systems by late 2020s. The scaling path: larger optical tweezer arrays, improved atom loading, fault-tolerant architectures.