Trump admits selling IBM’s stock was a mistake. Now he’s cheering its quantum future.
A burst of market candor and tech boosterism from Donald Trump has pushed IBM back into the spotlight: after saying he sold the stock too early—a mistake, in his telling—he’s now applauding the company’s push into quantum computing. Regardless of how you read the remark, the renewed attention is well-timed. Big Blue has spent the past few years rebuilding its narrative around hybrid cloud, AI, and a long-horizon bet on quantum—an area where IBM is both a scientific leader and a pragmatic strategist.
IBM’s comeback arc sets the stage
For much of the 2010s, IBM felt like a value trap: flat revenue, shrinking legacy businesses, and a sprawling portfolio that hid as much as it revealed. The company’s recent reset has been clearer. It spun off managed infrastructure into Kyndryl, focused investment around Red Hat and hybrid cloud, refreshed its mainframe franchise, and leaned into applied AI. The market has noticed, and investor sentiment has improved.
Quantum, however, is different from cloud or mainframes. It is not a near-term revenue engine. It’s a decade-scale technology race with high scientific risk, long R&D cycles, and uncertain commercial pacing. Yet if it delivers, the upside—new classes of simulations and optimizations beyond classical machines—could reshape industries. That’s the future Trump was cheering.
What IBM is actually building in quantum
IBM’s quantum strategy has three hallmarks: clear hardware roadmaps, a hybrid “quantum-centric” architecture, and an open software ecosystem.
– Hardware milestones: IBM has steadily scaled superconducting qubit counts while improving quality. It crossed the 100-qubit mark with its Eagle processor, moved to hundreds with Osprey, and surpassed a thousand qubits with Condor. In parallel, it introduced Heron, a lower-qubit, higher-fidelity design aimed at reducing errors—an acknowledgment that raw qubit counts mean little without stability. In late 2023, IBM began deploying Quantum System Two, a modular cryogenic platform designed to network multiple processors in the same system as the company moves toward larger, more reliable systems.
– Error reduction now, error correction later: Fully fault-tolerant quantum computing requires “logical” qubits protected by error-correcting codes, which in turn demand thousands of physical qubits per logical qubit—well beyond today’s machines. IBM’s near-term bet is error suppression and mitigation techniques that, when combined with better hardware, can deliver “quantum utility” on select problems sooner than full error correction will allow. The longer-term roadmap still targets error-corrected logical qubits later this decade.
– Hybrid, not standalone: IBM frames the future as quantum tightly coupled to classical HPC and AI. Through Qiskit and its cloud services, developers can stitch together classical pre- and post-processing with quantum circuits that run on real devices or high-fidelity simulators. That “quantum-centric supercomputing” model lowers the bar for practical experiments and makes IBM’s quantum hardware accessible alongside classical accelerators.
– An ecosystem beyond the lab: IBM has placed Quantum System One installations with research and enterprise partners in the United States, Europe, and Asia, and built the IBM Quantum Network that includes universities, labs, and companies. This seeding strategy aims to cultivate use cases in materials science, chemistry, finance, and logistics. The company has also been active in the transition to post-quantum cryptography; IBM researchers contributed to algorithms now being standardized by NIST, and IBM Consulting is helping clients plan crypto-agile upgrades.
Where quantum could matter first
Hype often races ahead of engineering reality, but several domains are plausible early beneficiaries as devices improve:
– Chemistry and materials: Quantum circuits that approximate molecular energies and reaction pathways could help design better catalysts, batteries, fertilizers, and specialty materials. Even small accuracy gains can compress R&D cycles.
– Optimization and scheduling: Hybrid quantum-classical techniques may offer speed or quality improvements in routing, portfolio construction, and manufacturing planning—especially when the cost of a slightly better answer is high.
– Machine learning: While general-purpose quantum ML is nascent, narrow tasks—feature mapping or kernel-based methods—could benefit as error rates fall.
Timelines remain uncertain. Expect incremental wins: proofs-of-advantage on narrow problems first, then broader classes as qubit fidelity and scale advance. For enterprise buyers, the practical path is exploratory: pilot projects today, competency building, and a plan to adopt as the technology crosses utility thresholds.
Policy tailwinds—and risks
Quantum sits at the intersection of national security and industrial policy. The United States, Europe, and Japan are funding quantum research, and export controls are shaping the competitive landscape. IBM’s footprint in New York state and its long-standing role in semiconductor R&D add to the political resonance of “American-made” advanced computing. On the security side, the advent of cryptographically relevant quantum machines—still years away—has already triggered a transition to quantum-resistant cryptography, creating nearer-term services revenue opportunities for companies like IBM.
The investor lens: what Trump’s enthusiasm does and doesn’t change
A high-profile endorsement changes the narrative, not the physics. The investment case for IBM’s quantum program hinges on execution against a long roadmap:
– Technical milestones: Lower two-qubit gate errors, stable mid-circuit measurement, scalable couplers between chips, and credible demonstrations of small error-corrected logical qubits.
– Software and workload maturity: Evidence that real customers can reproduce IBM’s “quantum utility” claims on meaningful problems, not just curated benchmarks.
– Ecosystem depth: Growth in the IBM Quantum Network, more System One and System Two deployments, and third-party toolchains that target IBM hardware via open standards.
– Transparency: Even if quantum remains non-material to revenue for several years, credible metrics—uptime, job throughput, error rates, number of active users and partner case studies—help investors separate signal from sizzle.
Competition is intense. Google pursues error-corrected surface codes; Microsoft emphasizes a cloud-first model with partner hardware; trapped-ion and photonic startups argue their qubits are more coherent, if harder to scale. No single architecture has “won.” IBM’s edge today is an integrated stack, a cadence of public milestones, and a large developer community.
Bottom line
If you sold IBM and watched it recover, regret is human. Cheering its quantum moonshot is easy; building it is hard. IBM’s quantum program is credible, measured, and globally influential, but still early. For technologists and investors alike, the right posture is disciplined optimism: track the engineering, pilot the use cases, and separate political theater from product reality. If IBM hits its next set of milestones—cleaner gates, modular scaling, and repeatable utility on real problems—the applause will have less to do with who’s cheering and more to do with what’s finally working.
What to watch next
– Independent replications of “quantum utility” results on IBM hardware
– Demonstrations of small, stable logical qubits and early error-corrected workflows
– Growth and outcomes from IBM Quantum Network industry pilots
– Clear migration paths for quantum-safe cryptography in large enterprises
– Evidence of modular, multi-chip quantum systems operating as a single coherent machine
