Why IBM's Quantum Computing Flywheel Is Just Warming Up

This article first appeared on GuruFocus.

IBM (NYSE:IBM) is currently at the inflection point, but not only due to nostalgia. There is no doubt that the company has created an opportunity to succeed in quantum computing.

IBM, after numerous experiments, is migrating to larger systems that can be used, and it has a roadmap to render the machines reliable. The company recently formed a partnership with AMD (NASDAQ:AMD), and constructed its own production facilities, which enabled it to produce hardware rapidly and repeatedly, much faster than many start-ups.

Is IBM fairly valued? Test your thesis with our free DCF calculator.

In addition to this, there is constant cash and profits through cloud and enterprise software, and IBM can invest in the long term without weakening the company. The adoption of IBM tools is evident, with most individuals employed in developing programmes and thousands of published papers.

The market picks up on the potential and simply believes that IBM is an outdated brand. Any fraction of the market that IBM takes of the quantum or AI hardware would be huge. IBM stands to expand alongside the development of quantum technology, and this article discusses why.

The Quantum Catalyst: A Road to Fault Tolerance

IBM seems to be all in on quantum computing. In December 2023, it introduced Condor, the largest quantum computer to date, with superconducting qubits. It will achieve quantum advantage by approximately 2026, and by 2029, it projects to have a reliable machine, named Starling. Starling will be designed with 100 million qubits and approximately 200 protected error-resistant logical qubits.

IBM will later release Quantum Blue Jay in the early 2030s, which will operate on approximately 2 million logical qubits (1 billion operations). In brief, IBM has a roadmap in place to scale to quantum advantage in the near future and has a massive system out by the close of the decade.

IBM leads in quantum. It operates two modular quantum systems at RIKEN in Japan. Our quantum cloud has approximately 600,000 users, and 300 organizations are clients. IBM will connect its quantum machines to traditional supercomputers such as RIKEN and will connect IBM System Two to Fugaku, which is among the fastest computers in the world.

IBM had invested heavily in a quantum pipeline. In April 2025, it declared a U.S. investment of $150 billion over five years, $30 billion in quantum and mainframe research.

Small but important initial results. IBM has made an estimated $1 billion in quantum since 2017, a small portion of its overall business totalling $65 billion; anyhow, it demonstrates that the idea has worked.

Story Continues

Raw quantum stocks typically contribute nothing to revenues until they become mainstream. The benefit here is that, although quantum sales are currently insignificant, any breakthrough in the AI, cloud, and consulting business of IBM can easily manage a quantum breakthrough without the risk involved. According to analysts, quantum revenue represents a minor fraction of the total earnings at IBM yet there is no significant threat and IBM can continue to move towards quantum.

IBM's practical advantages are concrete, not just promotional. The company leads in quantum patent filings (a larger count than most rivals), operates in-house fabrication and modular systems co-located with major supercomputers, and runs a broad IBM Quantum Network with hundreds of corporate and academic partners and roughly 600,000 users. That combination, intellectual property, engineering scale, partner integrations and large customer access, gives IBM a structural edge versus pure-play startups (which lack scale) and hyperscalers (which often focus on algorithms and cloud scale rather than end-to-end hardware and enterprise integration).

Overall, the comprehensive quantum roadmap offered by IBM and its partnerships with labs, governments, and companies form a force.

Assuming IBM makes these milestones quantum advantage by 2026 and reliable machine by 2029 IBM will change the future of IBM.

IBM/AMD - The Hybrid Supercomputer Partnership.

In August 2025, IBM declared plans to enter into an agreement with AMD and construct quantum-powered superbrained computers. The partnership combines the IBM quantum computing resources with the classical high-performance computing and neural network accelerators used by AMD.

Its idea is to develop hybrid systems that will allow tasks to be scheduled to both quantum and classic supercomputers, providing the best of both.

As one example, a molecular simulation would be processed using an IBM quantum processor and the data analyzed with an AMD-powered classical cluster.

Another example is specifically that IBM and AMD will incorporate the next generation of award-winning algorithms into IBM quantum machines using AMD CPUs, GPUs, and FPGAs. Quantum fault-tolerance could be enhanced by the aid of technology provided by AMD (to work with real-time error correction and AI workloads).

By the end of 2025, the partners are scheduled to demonstrate the first demonstration of IBM working hand-in-glove with AMD systems which demonstrates the quantum hardware. It is also an attempt to leverage, much like the AMD leadership in exascale computing. Currently, the two fastest supercomputers are powered by AMD EPYC CPUs and Accenture GPUs: Frontier and El Capitan computing systems.

In my opinion, the IBM-AMD alliance is a huge value addition. The larger market might reach the low-hundreds of billions in the coming decade, but the installed supercomputer business is significantly smaller now (quantum+HPC). It will be a long journey, Business revenues will arrive a long time, however, this demonstrates that IBM is also considering strategic partners to ensure that the company remains on the front line.

The Quantum Flywheel

IBM's vast core business fuels its quantum and AI ventures. The company generates roughly $6065 billion in annual sales from its hybrid cloud, software, services and mainframe businesses. From this, it ploughs about 1012% of revenue into R&D (including quantum and new technologies).

Thanks to this scale, IBM produces very high free cash flow, which is expected to be over $13.5 billion in free cash for 2025 and maintains strong margins. For instance, in Q2 2025, IBM's gross margin rose to 58.8% (GAAP) (vs. 56.8% a year earlier), illustrating that its AI/cloud growth is expanding profitability. In practice, this means the flywheel of IBM's business is quite tangible: its $65B core funds $68B of R&D each year, which in turn drives innovations (like new z16 mainframes, Red Hat enhancements, and quantum machines) that help IBM win more enterprise clients and licensing deals.

Today, roughly 45% of IBM's revenue comes from its AI and cloud segments, a proportion that will grow with Red Hat and generative AI. As this mix shifts, IBM's ROIC should improve: high-margin AI/cloud revenue is growing faster than legacy services, and the company's disciplined cost control means incremental investment in new tech is largely funded by existing cash flow. In short, IBM's stable cash engine can afford the long-term quantum gamble without straining its finances.

Financial Overview

The most recent earnings of IBM (Q2) were solid. The tech giant recorded an increase of 8% in revenues totalling $17 billion in 2025, with 5% increase in constant currency. Broad-based growth software (including Red Hat hybrid-cloud) increased by 10%, and infrastructure (servers/storage) increased by 14%. First in the list of mixing was slower at +3 with consulting (services), although higher-margin software and cloud services took the lead overall. Self-contained software mix increase and an augmentation of efficiency made IBM experience a 200-basis-point growth in gross margin (GAAP 58.8% versus 56.8% last year).Why IBM's Quantum Computing Flywheel Is Just Warming Up

GAAP net income on the bottom line amounted to $2.2 billion ($2.31 per share), which is approximately 20% higher than the bottom-line net income of $$1.83 billion a year ago. (IBM on a non-GAAP basis made earnings of $2.7B, or $2.80/share, as well, +17%.) The margin gains and the increase in revenue were driving the best profits.Why IBM's Quantum Computing Flywheel Is Just Warming Up

Moreover, IBM's free cash flow is a major strength: it created an approximate $2.8 billion in free cash during the second quarter (more than $2.6B one year ago).

Year-to-year, IBM has brought in operational cash and cash operating as well as $4.8 billion in free cash. This exchequer finances the dividend and new investments. The company moved back dividends paid in Q2, amounting to $1.6B, to shareholders, and increased guidance: the amount of free cash flow it will have in 2025 is now anticipated to be over $13.5 billion.

To conclude, IBM is profitable, expanding, and possesses a lot of cash. These cash flows can easily cover its $1.68 per share quarterly dividend (2.5% yield). The AI/Cloud book of business of the company is also running high: as Arvind Krishna reported, the IBM book of business kiosks of the new AI generated $7.5 billion at Q2, which shows that people are requesting the services of the new AI in IBM.

Moving forward, the next earnings that will be announced by IBM will be reported on October 22, 2025. The projected revenue is about $16.1billion and projected adjusted earnings per share are approximately $2.44.

Valuation looks cheap

IBM shares are cheap compared to conventional multiples in comparison with technological allies. IBM has also expressed a slower growth, exhibiting a trade at a much lower 24-25x forward earnings. Contrastingly, the forward P/E of Microsoft (MSFT) is stingingly poor: it is 33x, and even closer, Oracle (ORCL) shows more than 48x. Equally, the Price/Sales (3.9) and Price/Book (9) of IBM come nowhere near those of Microsoft and Oracle.

Rather than focus solely on the headline forward P/E of around 24 with 12% growth, we now emphasize risk-adjusted multiples and yields. On a free-cash-flow basis, IBM's forward EV/FCF is about 22, higher than Microsoft's roughly 18 but lower than Oracle's near 25. More importantly, IBM's 2.5% dividend yield stands well above Microsoft's 0.65% and Oracle's 0.7%. In effect, IBM trades at roughly the same EV/CF multiple as peers but with a much higher yield and far lower top-line growth, suggesting its valuation already reflects those risks.

This points to a risk-adjusted bargain, offering investors steady cash returns plus upside potential. IBM's EV/Sales ratio of only 2.8, less than one-third of Microsoft's 10.5 and below Oracle's 6.1, underscores the deep discount driven by legacy perceptions.

Given IBM's investment-grade balance sheet and consistent cash flow, its multiple appears justified if one accounts for slower growth, but any acceleration in AI or cloud traction could quickly re-rate the stock. Practically, this implies that re-rating at the present value of the earnings multiple of IBM stock can get the terms of stock originating at a much higher level.

As an example, expecting IBM to hit an EPS of about $12 in 2026, 22 coming in would push the stock into the downturn, high-250s (currently broken into the 265 level), and 30 would take it to the 360 level. Simply, even under those levels, IBM would continue trading below its counterparts.

Concerned investors note that the growth of IBM is low: the revenue growth is only +1.4% for the company in 2024, and consulting growth is relatively high (3% in Q2). However, in the event that the strategic transition of IBM to AI/cloud is successful, a small increment in growth would warrant additional multiple.

Notably, IBM has a strong income and dividends (approximately 2.5%2.7% yield). Even robust cloud growth like that of Oracle is yielding only at the mark of 0.7 per cent with a significantly higher PEG ratio. On a forward basis, therefore, IBM can be underpriced.

Investment Risks and Challenges

With this said, IBM is not risk-free. To begin with, business continues to be difficult: in FY2024, the increase in sales at IBM was reported as only 1.4%, which shows attitudes towards being extremely careful on the part of customers. The competition at IBM is quite intense in terms of AI and cloud computing. Megawinners in the market, such as Microsoft (MSFT), Amazon (AMZN) (AWS) and Google (GOOGL), secure immense clouds, and IBM should struggle to lure or regain enterprise customers.

In case hyperscalers drift, IBM's growth may come to a halt. Its consulting service/ segment also competes with Accenture, Deloitte, and others; hence, the timing of huge contracts can determine performance.

New initiatives also have the risk of execution. Both at the request level, both IBM and AMD/HPC gambits are exciting, and commercial payoff will probably have to wait several years. When such projects fail, it will not cause disaster to the core business of IBM (as is the case with pure-play stocks in quantum), but investors must not expect such projects to generate immediate revenue. Lastly, IBM is based in many countries, creating the risk of wars or currency fluctuations (e.g. technology export ban) deflating outcomes.

Conclusion

In my opinion, IBM offers a rare combination of yield today plus optional upside. At a 2.5% dividend yield and with >$13 billion annual free cash flow, investors have a lot of protection built in. At the same time, IBM's leadership in hybrid cloud/AI and its pioneering quantum roadmap provide long-dated optionality. In other words, shareholders effectively get quantum optionality without binary risk. Even if practical quantum breakthroughs take longer, IBM's core software and services business remains profitable and growing modestly, supporting the dividend and buybacks.

Conversely, if IBM executes on its AI/cloud and quantum initiatives (leveraging its 600K?user network, patent lead, and engineering scale), the stock could see significant upside surprise.

Even if quantum/computing breakthroughs take longer than hoped, IBM is still a profitable enterprise technology company with strong dividends and exposure to AI growth.

In simple terms, IBM is cheap for good reason, but that reason is already reflected in the price, meaning any further execution on AI/cloud/quantum could surprise to the upside.

IBM offers today's yield plus long-dated upside from AI/cloud and quantum, a risk-adjusted case for investors who want income now and credible optionality over the next several years.

View Comments