ACIS · QUANTUM COMPUTING RADAR · ISSUE 003 · 2026.09.06
System Usability Is
Replacing Qubit Count
Quantum Computing Research Radar · Issue 003
IBM has moved Nighthawk r2 from engineering release to real cloud access. Competition is shifting toward how much reliable computation can be completed per unit of time—an engineering upgrade, not yet a commercial-value breakthrough.
DIRECT ANSWER
Throughput may become quantum computing's next core metric
There was no new fault-tolerance or logical-qubit breakthrough this week. The real change is IBM Nighthawk r2 exceeding 100,000 circuits per second—up to 25 times Heron throughput—and opening to Premium, Flex and pay-as-you-go users. It improves system usability; it does not prove customer ROI or free cash flow.
Quantum is moving from who has more qubits to who completes more reliable computation per unit of time and cost.
01 · KEY CHANGES
Five signals that matter
01 · IBM Nighthawk r2｜From engineering release to real cloud access
- 120 programmable qubits
- 218 couplers and 120 independent reset units
- >100,000 circuits per second, up to 25× Heron
- About 25× lower initialization error
- Accurate results on circuits exceeding 7,500 gates
ibm_phoenix is now accessible through paid and usage-based plans. It reduces the time cost of repeated circuits, parameter sweeps and error mitigation, but adds no logical qubits and does not yet prove customer ROI. IBM ↗ · Platform update ↗
02 · Diraq × Equinix｜Entering the data center—with only eight qubits
The partners plan an eight-qubit silicon-spin deployment in Sydney in October, consuming under 20kW and co-locating with CPUs, GPUs and conventional servers. The test is remote operations, security, networking and data-center compatibility—not commercial-scale computation. Equinix ↗
03 · Xanadu｜1,000 logical qubits remain a roadmap target
The company targets up to 200 logical qubits by 2029 and more than 1,000 by 2031, backed by a signed C$195 million Canadian manufacturing commitment. The funding strengthens photonic fabrication, packaging and wafer testing; it does not make roadmap performance real today. Xanadu ↗ · Funding ↗
04 · Pasqal｜Protein encoding is an application experiment, not commercial success
Pasqal and True Nexus encoded structures related to protein gel formation on a neutral-atom system, but disclosed no comparative accuracy, speed, cost, laboratory discovery or customer payment versus classical methods. Pasqal ↗
05 · US Policy｜A unified commercial quantum policy remains proposed legislation
The American Quantum Competitiveness Act would put the Commerce Department in charge of commercial quantum technology, manufacturing and trusted supply chains, but it is not enacted or funded policy. GovInfo ↗
02 · EVIDENCE LADDER
Engineering advanced; commercial proof did not
Quantinuum's enterprise maturity framework likewise places most companies in exploration or experimentation, but it is customer education rather than new revenue evidence. Google, Microsoft, Rigetti and D-Wave offered no disclosure that changed the thesis. Quantinuum ↗
03 · MATURITY SCORE
Industry maturity rises to 48 / 100
| Platform | Score | Weekly | View |
|---|---|---|---|
| IBM | 85 | +2 | High throughput + real access |
| Quantinuum | 75 | — | Awaiting paid OCI use |
| IonQ | 73 | — | Revenue lead; segment clarity needed |
| D-Wave | 60 | — | Real use; volatile revenue |
| 60 | — | Science lead; limited access | |
| Xanadu | 59 | +1 | Manufacturing funded; roadmap unproven |
| Rigetti | 55 | — | Awaiting acceptance and payment |
| Pasqal | 54 | — | More applications; economics unproven |
| Microsoft | 52 | — | High potential; little external evidence |
| Diraq | 47 | +2 | Data-center validation milestone |
The one-point increase comes from system throughput and data-center deployability—not fault tolerance or profitability.
04 · CAPITAL ROTATION
The market is still pricing distant success
At the 4 September close: IONQ was $39.52, RGTI $15.20, QBTS $16.58, XNDU $9.72, while newly listed PSQL closed at $7.96 after falling 18.7% that day. As listed supply grows, capital is beginning to fragment across quantum equities.
Capital stage: high-valuation technology options, with material volatility, dilution and revenue-purity risk.
Emerging-Leader Radar
- 01IBM · UPGRADED
Delivering accessible system performance.
- 02Xanadu · UPGRADED WATCH
Manufacturing capital is real; the 2031 target needs stepwise validation.
- 03Diraq · PRIORITY WATCH
Stable operation in a normal data center matters more than eight qubits.
- 04IonQ · SEGMENT WATCH
Computing, communications, sensing and SkyWater contributions need separation.
05 · OPPORTUNITY & RISK
Opportunity lies in bottlenecks; risk lies in evidence mismatch
OPPORTUNITY
- Fast reset, real-time control and correction electronics
- Photonic packaging, wafer testing and module assembly
- Cryogenics integrated into commercial data centers
- Quantum + HPC + AI orchestration
- Independent validation, compilation and mitigation software
RISK
- Vendor benchmarks lack independent replication
- High throughput does not equal economic value
- Government funding confused with customer revenue
- Roadmaps treated as achieved performance
- New listings create valuation compression
06 · 90-DAY CATALYSTS
The next verifiable milestones
- 8 Sep | IonQ Investor Day: SkyWater integration cost, compute segmentation and system roadmap. IonQ ↗
- 10 Sep | IBM technical briefing: price-performance, utilization and external workloads.
- 13–18 Sep | IEEE Quantum Week: Quantinuum hardware, software and error-correction updates. IEEE ↗
- Whether Diraq completes the Equinix installation and opens external testing in October.
- Whether Oracle Helios enters preview and discloses utilization.
- Whether Rigetti's 108-qubit C-DAC system reaches acceptance and revenue recognition.
07 · RESEARCH ESCALATION SCAN
Two research escalations triggered
IBM Nighthawk r2 moves from hardware milestone to real cloud access. Engineering breakthrough / platform opening.
Validation memo: Why throughput may matter more than qubit count.
Near-term differentiation may first appear in reliable circuits completed per unit of time and cost.
08 · KEY QUESTIONS
Frequently asked questions
Has Nighthawk r2 achieved fault-tolerant quantum computing?
No. It improves physical-system throughput, reset speed and accessibility, but adds no new logical-qubit breakthrough this week.
Why might throughput matter more than qubit count?
Because parameter sweeps and error mitigation require repeated circuit runs. Higher throughput can reduce time and experimental cost, but result quality and customer ROI remain decisive.
Does Diraq entering Equinix prove commercialization?
No. It first tests data-center compatibility, remote operations and secure connectivity; the system has eight qubits and customer testing has not begun.
What stage are quantum equities in?
They are closer to highly valued, capital-market-funded technology options than mature computing platforms supported by free cash flow.
Source framework: company, government and market disclosures available through 6 September 2026. Maturity scores compare relative stages and are not investment advice.
IBM has crossed a system-usability milestone—not the fault-tolerance finish line. Engineering deserves an upgrade; revenue, customer ROI and free cash flow do not yet justify the same commercial upgrade.