Bernstein: Quantum Computing to Form CPU/GPU/QPU Hybrid Architecture; Infineon Rated Outperform
AI summary card
Bernstein: Quantum Computing to Form CPU/GPU/QPU Hybrid Architecture; Infineon Rated Outperform
The report argues that quantum computing will not replace classical hardware but will integrate into hybrid systems as specialized accelerators; superconducting, trapped-ion, and neutral atom technologies are the three mainstream paths, with quantum advantage expected around 2030.
- Future computing architecture will consist of CPUs, GPUs, and QPUs (Quantum Processing Units), with QPUs serving as specialized accelerators rather than replacements.
- Superconducting (led by IBM/Google), Trapped-Ion (led by IonQ/Infineon), and Neutral Atom are currently the three most advanced technology paths.
- Quantum advantage is expected to be achieved around 2030; prior to this, commercial value will be primarily limited to government and early-stage experimentation.
- Maintains 'Market-Perform' rating on IBM with a target price of $280; maintains 'Outperform' rating on Infineon with a target price of €74.
- Current pure-play public quantum computing companies imply a combined long-term market share of approximately 24%, with IonQ accounting for roughly 11%.
- The industry is not a winner-take-all market; different technology paths will serve specific application scenarios due to their distinct pros and cons.
Report interpretation
Overview
This report delves into the future landscape of quantum computing technology, focusing on analyzing different technology paths and key players. The core view is that quantum computing will represent the next significant evolution in computing but will not replace traditional CPUs and GPUs; instead, it will form a 'tri-processor' hybrid architecture consisting of CPUs, GPUs, and QPUs (Quantum Processing Units). The report evaluates six quantum technology modalities, identifying superconducting, trapped-ion, and neutral atom technologies as the three most competitive directions currently. Although quantum advantage is not expected until around 2030, the firm validated current valuations of listed companies using discounted cash flow and market share models, providing specific investment ratings for IBM and Infineon.
Core views
Evolution of Future Computing Architecture: The report posits that the future of advanced computing will be shaped jointly by CPUs, GPUs, and QPUs. In this hybrid quantum-classical system, QPUs will not replace classical hardware but will be tightly integrated as specialized accelerators. CPUs will handle system coordination and logic decomposition, GPUs will process massive parallel classical computations and prepare data for QPUs, while QPUs will tackle optimization problems and quantum simulation tasks that are intractable for classical systems due to exponential complexity. Giants like IBM and NVIDIA are driving this paradigm, exemplified by IBM's 'quantum-centric supercomputing' concept and NVIDIA's NVQLink architecture, aiming for seamless software-hardware synergy. Technology Path Competitive Landscape: The report analyzes six quantum technology modalities, identifying superconducting, trapped-ion, and neutral atom technologies as current leaders. Superconducting qubits are currently the most commercially mature and execution-ready technology, rapidly advanced by IBM and Google; advantages include fast gate speeds and strong scalability, though stability is lower and cryogenic environments are required. Trapped-ion technology offers the industry's highest precision and stability with long coherence times, making it the preferred choice for pure players like IonQ; Infineon is a leading foundry in this field, providing core hardware platforms; drawbacks include slower operation speeds and scaling challenges. Neutral atom architectures are flexible and easy to scale, but achieving high-fidelity gates at large scales remains challenging. Additionally, photonics, topological qubits, and silicon spin qubits remain in early or highly experimental stages; while promising, they are not yet mature. Quantum Advantage and Commercial Value Timeline: True economic value will be unlocked after achieving 'quantum advantage,' defined as when quantum systems complete commercially meaningful tasks that classical computers cannot feasibly match. Citing BCG data, the report estimates this inflection point will occur around 2030. Prior to this (pre-2030), provider annual revenue is expected to be approximately $1-2 billion, primarily from government labs and early corporate experiments; between 2030-2040, annual revenue could reach $15-30 billion as initial commercial use cases (e.g., drug discovery, materials modeling) emerge; post-2040, with the widespread adoption of fault-tolerant systems, annual revenue could reach $90-170 billion. Valuation and Implied Market Share: Given the early stage of quantum computing, the report uses reverse engineering to assess current valuations. Assuming a 2040 Total Addressable Market (TAM) of $90-170 billion (median $130 billion), the discounted present value to 2026 is approximately $26.6 billion. Assuming a 30% net margin for mature tech companies and a 28x SOX P/E multiple, the aggregate implied long-term market share for the six listed pure-play quantum companies is calculated at approximately 24%. Among them, IonQ, with the largest market cap, implies an 11% share, while others like Rigetti and Infleqtion have lower implied shares (4% and 2% respectively), suggesting significant upside potential if their technology paths succeed.
Analysis framework
The report employs a comprehensive analytical framework combining 'technology fundamentals + macro TAM estimation + reverse valuation derivation.' First, starting from technical principles, it compares the pros and cons of different qubit physical implementations (e.g., speed, stability, scalability) and combines this with major vendors' R&D progress to identify three mainstream technology paths. Second, at the macro level, it cites forecasts from authoritative institutions like BCG regarding the quantum computing TAM, distinguishing the pace of value release before and after 'quantum advantage' to emphasize a long-term perspective. Finally, regarding valuation, since most pure-play quantum companies are unprofitable, the report avoids traditional DCF models in favor of a 'Reverse Implied Market Share Method': using current market capitalization, assumed mature-stage valuation multiples (P/E 28x), and net margins (30%) to back-calculate the future revenue and market share implied by current pricing, comparing this against the discounted long-term TAM to determine which targets are relatively conservatively or aggressively valued. This approach avoids the immense uncertainty of directly forecasting cash flows for early-stage startups, focusing instead on the reasonableness of market expectations.
Methodology notes
Supply-side analysis of quantum computing technology paths
By comparing the physical characteristics of different technology paths such as superconducting and trapped-ion (e.g., gate speed, coherence time, manufacturing difficulty), the report analyzes which technology is better positioned for scalable supply, which is central to determining the industry's competitive landscape.
Reverse PE valuation based on maturity assumptions
For unprofitable early-stage tech companies, the report assumes they reach maturity by 2040, using industry-average P/E multiples (28x) and net margins (30%) to back-calculate implied future earnings and revenue from current market caps, assessing whether current stock prices have priced in future growth.
Quantum Advantage as an industry inflection point
The report defines 'Quantum Advantage' as the critical inflection point where the industry shifts from R&D-driven to commercial monetization (expected 2030); market demand is limited before this point, and large-scale enterprise applications will only emerge afterwards. This phased segmentation forms the basis for forecasting revenue growth.
Differentiated competitive advantages of technology paths
The report notes that different technology paths (e.g., superconducting speed vs. trapped-ion stability) have distinct trade-offs, making a 'winner-take-all' outcome unlikely; instead, multiple architectures will coexist, each occupying specific application scenarios, which impacts the long-term market position of respective companies.
Asset mapping & comparison
Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).
- IBM (IBM)Leader in superconducting quantum computing, offering full-stack solutions
- Strengths
- Possesses the most comprehensive roadmap, strong hardware (Eagle/Heron chips) and software (Qiskit) ecosystem, driving hybrid architecture standards.
- Weaknesses
- Quantum business accounts for a small proportion, subject to high news-flow driven volatility; uncertainty regarding AI small model strategy.
- Comparison
- Larger scale and stronger risk resistance compared to pure players, but lower quantum purity.
- Risks
- Weakness in consulting business; valuation pressure if quantum progress falls short of expectations.
- Infineon (IFX.GR)Key hardware foundry for trapped-ion quantum computing
- Strengths
- Leading trapped-ion foundry, supplying QPUs to top players like IonQ; robust semiconductor manufacturing foundation, reliable 3rd generation ion trap technology.
- Weaknesses
- Volatility in traditional automotive semiconductor demand may impact overall performance.
- Comparison
- Unlike full-stack companies, Infineon focuses on underlying hardware enablement, with a business model closer to traditional semiconductors.
- Risks
- Slowing EV penetration; SiC oversupply; valuation multiple compression.
- IonQ (IONQ)Largest pure-play quantum computing company by market cap, focused on trapped-ion
- Strengths
- Ultra-high fidelity, all-to-all connectivity, and long coherence times; plans to launch Tempo system in 2026.
- Weaknesses
- Revenue scale still small (2026E $260M), not yet profitable.
- Comparison
- Leading position in trapped-ion sector, 11% implied market share, relatively high valuation.
- Risks
- Technical scaling difficulties; commercialization slower than expected.
- Rigetti Computing (RGTI)Pure player in superconducting quantum computing
- Strengths
- Full-stack approach, modular processor design improving error mitigation; implied market share of only 4%, relatively low valuation.
- Weaknesses
- Small revenue scale (2026E $24M), intense competition.
- Comparison
- Shares superconducting path with IBM/Google, but offers greater elasticity as a pure player.
- Risks
- Technical bottlenecks in superconducting path; high cash burn rate.
- Infleqtion (INFQ)Leader in neutral atom quantum technology
- Strengths
- Diverse product line (computers, clocks, sensors); implied market share of only 2%, attractive valuation.
- Weaknesses
- Neutral atom technology still faces challenges in large-scale high fidelity.
- Comparison
- More diversified business compared to single-focus computing companies, reducing reliance on a single application.
- Risks
- Neutral atom technology path fails to gain mainstream adoption.
Key data
- 2040 Quantum Computing TAM Forecast$90B - $170BBCG estimated range for provider annual revenue, median $130B
- Expected Timing for Quantum AdvantageAround 2030Point in time when commercially meaningful superiority over classical computers is achieved
- 2026 Discounted Quantum TAM~$26.6BMedian 2040 TAM discounted to 2026 at a 12% discount rate
- Implied Aggregate Market Share of 6 Pure-Play Companies24%Derived based on current market caps and assumed mature-stage margins/valuation multiples
- IonQ Implied Market Share~11%Highest among the 6 pure players
- IBM Target PriceUSD 280Based on 20x P/E applied to 2027E EPS of $13.93
- Infineon Target PriceEUR 74Based on 25x P/E applied to Q5-Q8E EPS of €2.93
Impact & implications
For the industry, quantum computing will undergo a long period of technological iteration and will not disrupt classical computing in the short term but serve as a complement. For investors, this implies a need to focus on companies with leading positions in specific technology paths whose valuations do not fully reflect long-term potential. The report specifically notes that Rigetti (superconducting) and Infleqtion (neutral atom) have low implied market shares based on current market caps (4% and 2% respectively); if their technology paths ultimately prove competitive, their stock prices may have significant room for re-rating. Conversely, while tech giants like IBM and Google possess resources, the contribution of their quantum businesses to overall valuation requires time to verify. Infineon, as a key hardware supplier in the trapped-ion space, benefits from the high stability advantages of this path and has received an Outperform rating.
Risks
- Historical stock prices of pure quantum players are highly driven by news flow; as quantum businesses grow as a percentage of revenue for large companies like IBM, these majors may also become more volatile.
- Uncertainty regarding IBM's AI small model strategy could pose upside or downside risks.
- Weak bookings in IBM's consulting business may signal broader business weakness.
- Infineon faces risks from slowing automotive demand, declining EV penetration, price erosion in SiC solutions, and oversupply.
- Overall weak semiconductor demand could lead to valuation multiple compression for Infineon.
- Delay in achieving quantum advantage, resulting in slower-than-expected commercialization.
- Technology path failure risk: Currently leading superconducting, trapped-ion, or neutral atom technologies could be overtaken by other early-stage technologies (e.g., topological, photonic).
What to watch
- Progress toward achieving Quantum Advantage, particularly key technical milestones before 2030.
- Breakthroughs in error correction capabilities and scalability across different technology paths (superconducting, trapped-ion, neutral atom).
- Revenue growth and loss narrowing of pure-play quantum computing companies to validate commercialization capabilities.
- Progress by big tech companies (IBM, Google, Microsoft) in establishing hybrid architecture standards and ecosystems.
- Order intake for Infineon's trapped-ion foundry business and its resilience against the broader semiconductor cycle.