Diraq Quantum Roadmap | Scaling to Millions of Qubits

Mapping the Path to Millions of Qubits

Diraq, a Sydney-based quantum hardware firm, has released a white paper titled The Case for Silicon. This document outlines the company strategy for achieving utility-scale quantum computing. Their primary goal involves placing millions of spin qubits on a single silicon chip. This approach pivots away from exotic materials and relies on existing CMOS-compatible semiconductor manufacturing processes.

Andrew Dzurak, the founder and CEO of Diraq, argues that previous industry roadblocks focused too narrowly on singular physics breakthroughs. He suggests that building a viable computer requires balancing physics, engineering, and economics at once. Silicon serves as the anchor for this strategy because it benefits from decades of established semiconductor infrastructure. By using existing foundries, Diraq expects to reach a milestone of 150,000 physical qubits and 1,000 logical qubits by 2029.

Technical Milestones and Hardware Projections

The roadmap described in the white paper moves quickly after the initial 2029 target. By 2031, the company anticipates scaling to more than 2 million physical qubits and over 10,000 logical qubits. These benchmarks depend on the integration of quantum processors into standard rack-scale cryogenic systems. Such designs allow these machines to fit directly into modern data center environments.

Diraq defines four pillars for its architecture. The system must be scalable to millions of qubits. It must remain economical, with a projected cost of under $1 per qubit. The design requires full deployment capability into current infrastructure. Finally, the hardware must perform fast, error-corrected operations. The company aims for one million error-corrected operations per minute, a speed they claim is essential for true commercial utility.

Integration and Future Industry Standing

The reliance on CMOS manufacturing is the defining feature of Diraq's technical direction. Because this fabrication technology already produces billions of standard computer chips annually, it creates a cost-effective pathway that custom quantum platforms often lack. This choice allows the firm to avoid the expenses associated with building proprietary manufacturing facilities from the ground up.

Diraq has established a physical presence in the United States to match its Sydney roots, with locations in Palo Alto, Chicago, and Santa Monica. The company is currently engaged with high-profile industry partners such as NVIDIA, Dell Technologies, and GlobalFoundries. They are also one of 11 organizations participating in Stage B of the DARPA Quantum Benchmarking Initiative.

The broader industry context involves a pivot toward reliability and error correction. While early quantum development focused on merely keeping qubits stable, the current phase demands systems that can actually run complex algorithms without crashing. Diraq’s focus on integrating with existing data center cooling and logic hardware suggests a shift toward practical utility rather than experimental proof of concept. The success of this path will depend on whether silicon spin qubits can maintain coherence at the scale required for the millions-level targets they have publicly set. If they succeed, the barriers to entry for quantum-powered applications in fields like chemistry and materials science may drop significantly by the start of the next decade.