Quantum Hardware — Hyperscaler / Diversified Embedded

CGI has signed a partnership with D-Wave Quantum Inc. that folds D-Wave's hardware and hybrid solvers into CGI's client-facing technology services, moving the company's quantum computing efforts from slideware to a signed agreement. The deal lands as CGI's momentum has been mixed: its CA$101.39 share price sits after a 7 day share price return of 5.72% and a 90 day share price return of 5.90%, while its year to date share price return and 1 year total shareholder return are both down close to 20%. On the most widely followed view of the stock, a fair value of about CA$119.85 sits above the current CA$101.39 share price, putting the new D-Wave partnership into a context of already modelled AI and modernization upside. CGI's heavy reliance on large government and enterprise contracts, combined with integration risk from recent acquisitions, could weaken that upbeat, AI-driven valuation case.

Bank of America initiated coverage of IonQ on September 28 with a Buy rating and a $60 price target, calling the company the largest public pure-play quantum company by revenue. The call lands as IonQ reported $80.1 million of Q2 2026 revenue, up 287% year-over-year, its strongest quarter on record and a fifth consecutive quarter of record results, with organic revenue up 132%. Management raised full-year 2026 IonQ revenue guidance to $280 million to $290 million and maintained its expectation for approximately 100% organic growth, while remaining performance obligations reached $485 million, up from $122 million a year earlier. Following its $1.8 billion acquisition of SkyWater, IonQ said it received its first fully integrated quantum processing units from the foundry and plans to begin commissioning 256-qubit systems in 2027, and it announced Superion 256 deployments at Florida International University and NVIDIA's Accelerated Quantum Research Center. BofA also pointed to IonQ's semiconductor-enabled approach to scaling qubit counts and its expansion across quantum computing, communications and sensing.

IonQ said on September 23 that its Superion 256 will become the first on-premise quantum processor deployed at Nvidia's Accelerated Quantum Research Center, with the deployment planned for 2027. Nvidia will place IonQ's quantum processing unit alongside conventional accelerated computing hardware in a setting designed to develop large-scale hybrid systems, targeting applications including financial optimization, materials science and computational chemistry. IonQ introduced Superion 256 on September 8 and has begun accepting orders, with first customer deliveries expected in 2027 and commercially manufacturable fault-tolerant systems targeted for 2028. On September 22, IonQ reported a decoder breakthrough, demonstrating a real-time decoder running continuously on a single conventional CPU across benchmark circuits simulating up to 408 logical qubits and more than 31.5 million quantum operations, with as little as 0.02% additional execution time. The financial value of the Nvidia deployment has not been disclosed, and IonQ's 2026 revenue guidance of $450 million to $460 million includes SkyWater revenue from July 31 through year-end, so the increase cannot be attributed entirely to accelerating organic quantum-computing revenue.

International Business Machines Corporation has strengthened its decades-long partnership with Marist University through an Innovation Incubator focused on artificial intelligence research, enterprise computing and student career readiness. The program will use the IBM z17 mainframe to support initial projects in AI, business, finance and civic engagement, including autonomous AI agents in controlled marketing campaigns, quantum optimization algorithms for analyzing real-time financial data and improving portfolio allocations, and ways AI can improve polling methods, survey design and models of voter participation. IBM experts will work with Marist students and faculty on research addressing business, societal and community challenges, with applications extending into business, communications, political science and environmental studies. IBM faces competition from Microsoft Corporation, which is advancing AI through its Copilot products and intelligent agents, and from Alphabet Inc., which is expanding AI through new Gemini models used in healthcare, weather forecasting, education and scientific research. IBM shares have lost 13% over the past year against the industry's growth of 212.1%, and the stock trades at a forward price-to-sales ratio of 3.02 versus the industry average of 5.29. Earnings estimates for 2026 have declined 0.1% to $12.33 over the past 60 days, while estimates for 2027 have decreased 0.2% to $13.22, and IBM currently carries a Zacks Rank #4 (Sell).

Hewlett Packard Enterprise said it will integrate QuEra Computing's fault-tolerant quantum systems with its Cray supercomputers for on premises high performance computing environments. The collaboration targets national labs, enterprises and high performance computing centers that want quantum and classical workloads running on a single infrastructure. QuEra's quantum technology is being positioned for scientific and energy research tasks that run alongside traditional Cray supercomputing jobs inside existing data centers. The move fits HPE's wider push to serve complex on premises workloads for governments and large enterprises, reinforcing its role as a full stack provider rather than just a server vendor. The real test will come from the first disclosed integrated testbeds or customer deployments where QuEra systems sit alongside HPE Cray supercomputers, especially at national labs or major HPC centers, along with any concrete timelines or contract details.

IBM has placed its z17 mainframe at Marist University's new Innovation Incubator, sending shares up about 2.2% to $236.55 Wednesday morning from Tuesday's $231.38 close. Marist plans to test autonomous marketing agents and explore AI applications in polling, and researchers are also scoping a project that would use quantum optimization to build portfolio allocations. IBM provided the machine for students and faculty to develop and test applications, and its announcement gives no dollar value for the system or revenue forecast for the collaboration. The stock still sat 2.59% below its $242.84 GF Value estimate, a gap that gives investors a valuation reference but that the campus project alone does not close. The investor case here takes patience, since getting future developers comfortable with IBM's technology could pay off if their work reaches enterprise customers, but the z17 at Marist shows where IBM wants its platform used rather than what customers will pay for the applications that emerge.

IonQ announced it successfully developed and tested what it calls the industry's first end-to-end real-time quantum error correction decoder running on a single standard off-the-shelf central processing unit, sending its shares up 12% in morning trading. The news lifted the broader quantum computing trade, with QTUM, the largest ETF tracking the sector at about $5.7 billion in assets, up 41% year to date. IonQ was the top trending ticker on the Yahoo Finance platform following the announcement. The decoder addresses one of quantum computing's central obstacles, the errors that make current systems unsuitable for commercialization, and is seen as a significant step forward for the industry. The discussion also noted that IBM has made quantum a major part of its technological push and plans to spin off its quantum unit, while government investment in the field continues.

Intel CEO Lip-Bu Tan said at Splunk's .conf26 conference that CPUs, GPUs and quantum processors will likely work together across different applications, citing error correction as an area where the technology still needs to advance. Days later, IonQ said it had tested a real-time quantum error-correction decoder running on a single standard CPU, with the system tested on simulated workloads representing up to 408 logical qubits and more than 31.5 million operations. IonQ said the decoder added as little as 0.02% to execution time under standard operating-noise assumptions, though the research used simulated workloads rather than a quantum computer operating with 408 logical qubits, so it does not establish that the technology is ready for commercial-scale deployment. Still, the test illustrates how conventional computing could handle error correction while quantum processors perform specialized calculations. INTC shares rose 1.7% Tuesday to $123.86, while IonQ has drawn renewed attention from investors following the milestone, and the next catalyst is whether IonQ can demonstrate the decoder on increasingly large, real quantum systems.

Photonic Inc. and Microsoft are collaborating to advance quantum resource estimation for future large-scale quantum computing systems, the companies announced in a joint press release. The effort builds on their existing strategic partnership, combining Photonic's expertise in next generation error correction codes and distributed quantum computing with Microsoft's Quantum Resource Estimation framework. The companies are exploring how architectural choices, networking requirements between quantum systems, and emerging error-correction approaches influence the resources required to run quantum algorithms at scale, including qubit counts, run time, and overall system overhead. Photonic's Quantum Low-Density Parity Check error-correction code family SHYPS is cited as an important advance in reducing the overhead required to operate large-scale quantum computers by helping systems do more with fewer physical qubits. Paul Terry, Chief Engineering Officer at Photonic, said the partnership has co-innovation at its core and aims to give the industry a clearer picture of what it will take to run practical applications on tomorrow's quantum hardware, while Matthias Troyer, Technical Fellow and CVP at Microsoft Quantum, said the work will explore how distributed architectures and error-correction advances influence future scalability and performance. In advance of the release of these capabilities on the Microsoft Quantum Resource Estimator, Photonic is inviting organizations to sign up for updates on the infrastructure, resources, and architectural considerations required to realize value from quantum computing at scale.

B. Riley Securities is maintaining bullish ratings on IonQ, D-Wave Quantum and Rigetti Computing ahead of Quantum World Congress, betting that a shift in government policy from pure research toward commercialization could open a more meaningful revenue path for the quantum-computing sector. The firm kept Buy ratings with price targets of $100 for IonQ, $40 for D-Wave and $35 for Rigetti. Analyst Craig Ellis pointed to the Department of Energy's From Lab to Market session at the September 23-25 conference as a sign that federal policy may increasingly focus on moving quantum technologies into commercial applications. The event will also feature a two-hour National Quantum Updates session with more than 10 governments, participation from the White House Office of Science and Technology Policy, and back-to-back keynotes from Microsoft and IBM. B. Riley's thesis is sector-wide, but the report provides no fresh revenue forecasts, cash-flow estimates, technical milestones or new valuation work behind the three price targets, leaving the stocks dependent on long-term expectations rather than near-term financial evidence.

IBM raised its quarterly dividend by a single penny, from $1.68 to $1.69, in a declaration dated April 22, 2026, marking its 31st consecutive year of increases but adding only about four cents to the annualized payout. The one-cent step continues a run of five straight penny bumps, a sharp slowdown from 2015, when the quarterly payout rose from $1.10 to $1.30. IBM's dividend remains well covered: the company produced $14.73 billion in free cash flow in 2025 against a payout of $6.255 billion, and CFO Jim Kavanaugh told analysts on the Q2 call that IBM returned $3.2 billion to shareholders in dividends through the first half of the year. The cash is increasingly being redirected toward acquisitions of Confluent and HashiCorp, $62 billion in debt, and more than $10 billion committed to quantum computing over five years, while CEO Arvind Krishna pointed to a generative AI book of business exceeding $12.5 billion. IBM offers the highest yield among its peers at 2.83% but the slowest growth, trailing Microsoft, which pushed its quarterly dividend from $0.91 to $0.98 in September on a run from $0.13 in 2010, as well as Cisco Systems and Oracle. With IBM shares down 20.71% year to date at $230.17, the total-return case now rests on the AI reinvestment thesis rather than dividend growth.

Mitsubishi Electric Corporation announced that its ME Innovation Fund has invested in OptQC Corp., a Japanese startup developing optical quantum computer hardware, marking the fund's sixteenth investment to date. OptQC originated from the University of Tokyo and was established on technologies developed by the university's Furusawa Laboratory, a world leader in optical quantum computing. The startup's optical quantum computer uses a proprietary architecture designed to limit increases in system size as qubit counts grow, and it has already delivered its first commercial system to a public research institution as well as hardware modules to private-sector companies. Through the investment, Mitsubishi Electric expects to gain early insights into OptQC's hardware development technologies, related components, customers and market trends. The company aims to lead the field of quantum computing and create new business opportunities by applying those insights to its own quantum technologies, software implementation knowhow and application development in industrial fields.

Nvidia unveiled new quantum-development software Monday, claiming a sevenfold acceleration in a research workflow, even as its shares traded around $210.87, down approximately 3.5%. The company said Fermilab compressed five months of research work into three weeks in an architecture-development workflow, and that its new open-source CUDA-Q Logical layer supports development of fault-tolerant quantum applications. The distinction matters because researchers completed a design workflow faster, which does not establish faster quantum hardware, commercial viability or additional sales. For shareholders, the potential payoff comes from supplying the computing resources behind quantum research, though that commercial outcome remains unproven. GuruFocus shows a GF Score of 95/100, with profitability and growth standing out while a weaker GF Value reading puts valuation under greater scrutiny.

NVIDIA announced an expansion of its CUDA-Q open source platform with CUDA-Q Logical, an orchestration layer that provides a programmable, verifiable approach to developing applications for fault-tolerant quantum computers. Fault-tolerant quantum processors with logical qubits are essential for useful quantum computing, and CUDA-Q Logical lets researchers design and orchestrate the many components such applications require, switching between algorithm, error-correction and hardware options to find optimal configurations. Fermilab used CUDA-Q Logical to validate prior results and evaluate physical qubits, runtimes and other resource requirements across different error-correction approaches and quantum hardware, accelerating fault-tolerant algorithm development from five months to three weeks, a 7x speedup. Iceberg Quantum used CUDA-Q Logical to model its fault-tolerant architecture for Diraq's qubits, showing how 1,000 logical qubits can be created with just 150,000 physical qubits, roughly 10x fewer than Diraq's previous estimates. Sandia National Laboratories developed QUOPS, a new independent cross-platform benchmark measuring progress toward utility-scale quantum applications, and a QUOPS reference implementation is now available in NVIDIA CUDA-Q; Sandia shared early results in a preprint ahead of IEEE Quantum Week, reporting initial QUOPS benchmarks for QPUs from Google, IBM and Quantinuum. CUDA-Q Logical is already being used by QPU makers and labs including Fermi National Accelerator Laboratory, Infleqtion, IQM Quantum Computers, QCDesign Quantum Motion and Sandia National Laboratories, and is now available through GitHub.

US blockchain company Eigen Labs published research on the paper preprint site arXiv on September 10 showing that the amount of computation needed to break ECDSA, the cryptographic scheme used by Bitcoin and Ethereum, with a quantum computer is less than previously assumed. The research team optimized the "point addition" operation, which is unavoidable in cryptographic cracking calculations, so that it can be executed with 1,151 logical qubits, the basic unit close to a quantum computer's working memory, and also held the number of "Toffoli gate" operations, among the heaviest computational loads, to about 1.3 million. The resource metric combining qubit count and operation count came in more than 50% lower than the circuit value presented in March 2026 by Google Quantum AI, Google's quantum computing division, but the two differ in calculation methods and conditions, and the paper states that it "does not prove superiority or inferiority and is limited to a numerical comparison." More than 100 humans and AI agents took part in this research starting in May 2026, and in June the verification mechanism and scoreboard were made public so that anyone could submit improvement proposals. However, no quantum computer exists that can actually run this circuit, and the number of logical qubits IBM aims for by 2029 is said to be only 200.

Diraq and Equinix announced plans to deploy a Diraq quantum computer at an Equinix data center in Sydney, Australia, marking the world's first silicon spin quantum computer to operate in a shared commercial data center. The system features an eight-qubit silicon chip with self-contained cryogenic cooling and control electronics, drawing less than 20kW of power and fitting within existing data center infrastructure. Scaling to higher qubit counts requires only a chip replacement, making the system easily upgradable. The collaboration aims to test real-world performance, chart a path to commercial scale, and begin partner and customer conversations. Installation at Equinix's Sydney data center is expected to be completed in October 2026.

Rigetti Computing reported that its first two 9-qubit systems delivered in 2026 went to commercial customers, marking a notable shift from a year or two ago when commercial demand was largely absent. Management attributed a significant portion of first-half revenue growth to commercial organizations purchasing on-premises quantum systems, while the broader pipeline now spans commercial enterprises, universities, and governments. The company's $8.4 million C-DAC order for a 108-qubit system remains on track for fourth-quarter revenue recognition, and the HPE-Pittsburgh Supercomputing Center collaboration is expected to place a full 9-qubit Rigetti system into an HPC environment in 2027. D-Wave Quantum posted $3.1 million in second-quarter 2026 revenues, essentially flat year over year, with commercial enterprises accounting for roughly 62.4% of revenues, up from 45.1% a year earlier. IonQ completed its acquisition of SkyWater Technology, creating a U.S.-based quantum foundry and strengthening its control over semiconductor manufacturing, advanced packaging, and supply-chain capabilities.

IonQ reported second quarter 2026 revenue of $80.1 million, up 287% year over year and 20% above its own expectations, driven by deployments of its fifth-generation quantum computing systems. The company also raised its full-year revenue guidance to a range of $280 million to $290 million, reflecting strong first-half performance. IonQ completed its $1.8 billion acquisition of SkyWater Technology on July 31, 2026, creating a vertically integrated full-stack quantum platform company, and received its first fully featured quantum processing units from SkyWater, which are now undergoing testing. Remaining performance obligations reached $485 million, up from $122 million a year earlier, while GAAP net loss was $1.9 billion, including a $1.6 billion non-cash warrant valuation charge. The company ended the quarter with $3.0 billion in cash and investments, and expects to begin commissioning its 256-qubit systems in 2027.

Quantinuum CEO Rajeeb Hazra announced an industry-first partnership with Oracle to integrate its Helios quantum computer into Oracle's data center. The deal will place Helios, described as the world's most accurate quantum computer, into Oracle's compute infrastructure and offer it as a service. This integration aims to let Oracle customers deploy hybrid workloads combining AI and quantum computing seamlessly, without managing additional hardware.