Steven GellerQuantum Computing, End to End

Book contents

Current section

Part IX. Company and Investing

  1. The Quantum Company Landscape
  2. Hardware Modality Diligence
  3. Software and Developer-Platform Diligence
  4. Error-Correction-Stack Diligence
  5. Application-Company Diligence
  6. Supply-Chain and Infrastructure Diligence
  7. Market Timing and Wedge Selection
  8. Moats, Partnerships, and Standards
  9. Reading Roadmaps Like an Operator
  10. Investment Memo and Startup Thesis Templates

Part IX. Company and Investing · Chapter 73

The Quantum Company Landscape

A hardware startup and a post-quantum security vendor are both "quantum companies," and comparing them on one metric is how diligence goes wrong. This chapter segments the field into layers and gives each layer its own proof gate.

Artifact
In this chapter 9 sections

Map companies by the layer they control, the interfaces they depend on, the buyer and budget they serve, and the evidence behind present revenue; compare only peers with the same dependency and economic boundary.

A company list becomes analytical only after it exposes interfaces and dependencies. A hardware manufacturer, compiler vendor, application consultancy, cryogenic supplier, and developer platform sell to different buyers on different budgets. Ranking them on one score conceals rather than resolves those differences.

A company map begins with interfaces

The map uses nodes for controlled layers and edges for necessary access, standards, suppliers, or buyer systems. Each node records buyer, payment evidence type, delivery unit, upstream concentration, and interface. A peer group exists only when those fields are commensurate. The graph’s important output is correlated exposure, not a colorful taxonomy.

Dated stack map: layers, buyers, dependencies

The full-stack review supports the layered architecture. GAO provides a dated category assessment. OpenQASM, Qiskit, and Cirq identify current software interfaces without proving portability. ACM and randomized benchmarking supply inspectable artifact and device-evidence concepts. The benchmarking review cautions against collapsing incomparable tasks into one ranking.

The Quantum Company Landscape: claim and source ledger, frozen 14 August 2026
Evidence IDRecordsClaim supported
E-73-1Lieven M. K. Vandersypen et al., A look at the full stack (2021)
OpenQASM Technical Steering Committee, OpenQASM 3 specification (2026)
Quantum systems comprise application, software, compilation, control, device, and infrastructure layers with material interfaces between them.
E-73-2U.S. Government Accountability Office, Quantum Computing and Communications: Status and Prospects (2021)Government assessments distinguish technology categories and maturity rather than treating quantum firms as one homogeneous market.
E-73-3OpenQASM Technical Steering Committee, OpenQASM 3 specification (2026)
IBM Quantum, Qiskit documentation (2026)
Google Quantum AI, Cirq documentation (2026)
Open languages and official SDK interfaces are evidence for portability claims only to the extent actually tested.
E-73-4Association for Computing Machinery, Artifact Review and Badging (2026)
Easwar Magesan, J. M. Gambetta, and Joseph Emerson, Scalable and robust randomized benchmarking of quantum processes (2011)
Timothy Proctor et al., Benchmarking quantum computers (2025)
Reproducibility and benchmark policy provide inspectable evidence fields for technical claims.

Revenue evidence must survive category changes

The anonymized map places two application firms and a workflow platform on the same hardware-access dependency. It separately places a control supplier behind two hardware nodes. A portfolio that appears diversified by company count therefore has two concentration points: backend access and hardware production volume. Those edges matter more than the labels printed in a market map.

Artifact contract. A dated node-edge table and executable cycle/concentration analyser. Every company category has a buyer and upstream dependency; the analyser reports shared critical dependencies and forbids cross-category score ranking.

Executable reference fixture
NODE_FIELDS = {"name": str, "layer": str, "buyer": str, "dependency": str, "source_id": str}
def inspect_map(dossier):
    errors = []
    threshold = dossier.get("concentration_threshold", {})
    if type(threshold.get("value")) is not int or threshold.get("unit") != "companies":
        errors.append("threshold_schema")
    counts = {}
    groups = {}
    for node in dossier.get("nodes", []):
        missing = [name for name, kind in NODE_FIELDS.items()
                   if name not in node or type(node[name]) is not kind or not node[name]]
        if missing:
            errors.append(node.get("name", "node") + ":" + ",".join(missing))
            continue
        counts[node["dependency"]] = counts.get(node["dependency"], 0) + 1
        groups.setdefault((node["layer"], node["buyer"]), []).append(node["name"])
    limit = threshold.get("value", 10 ** 9)
    concentrations = sorted(name for name, count in counts.items() if count >= limit)
    return {"decision": "invalid" if errors else ("concentrated" if concentrations else "diversified"),
            "errors": errors, "concentrations": concentrations,
            "comparison_groups": {" / ".join(key): sorted(value) for key, value in groups.items()}}
dossier = {"scenario": "portfolio dependency map", "concentration_threshold": {"value": 2, "unit": "companies"},
           "nodes": [
               {"name": "application-a", "layer": "application", "buyer": "industrial R&D",
                "dependency": "backend-access", "source_id": "scenario:company-a filing"},
               {"name": "application-b", "layer": "application", "buyer": "operations",
                "dependency": "backend-access", "source_id": "scenario:company-b filing"},
               {"name": "hardware-c", "layer": "hardware", "buyer": "HPC",
                "dependency": "control-supplier", "source_id": "scenario:company-c filing"},
               {"name": "hardware-d", "layer": "hardware", "buyer": "research",
                "dependency": "control-supplier", "source_id": "scenario:company-d filing"}]}
base_result = inspect_map(dossier)
bad_nodes = [{key: value for key, value in dossier["nodes"][0].items() if key != "buyer"}, *dossier["nodes"][1:]]
bad_result = inspect_map({**dossier, "nodes": bad_nodes})
strict_result = inspect_map({**dossier, "concentration_threshold": {"value": 3, "unit": "companies"}})
assert base_result["concentrations"] == ["backend-access", "control-supplier"]
assert bad_result["decision"] == "invalid" and "buyer" in bad_result["errors"][0]
assert strict_result["decision"] == "diversified" and strict_result["concentrations"] == []
print(f"PASS: 73 stack map bottlenecks={base_result['concentrations']} invalid={bad_result['errors']} threshold3={strict_result['decision']}")
Quantum stack dependency graph: inspected record
NodeControlled layerBuyerCritical upstream edge
Application Aworkflow and validationindustrial R&Dbackend access
Application Boptimization serviceoperations teambackend access
Control supplierreadout electronicshardware builderproduction volume
Hardware Cdevice and controlresearch/HPCcontrol supplier

Exact validation command: python3 tools/validate_briefs.py --briefs data/editorial_briefs_64_87.json --from 64 --through 87 --check-rewritten-sources --execute-artifacts

A dependency graph exposes correlated bets

Use the map to choose questions. For a concentrated backend edge, request contracts, portability tests, queue and pricing history, and the failure plan. For a supplier edge, request qualified alternatives, lead times, capacity, and unit demand. Do not convert the graph into a total attractiveness score; company-specific economics and evidence belong in their own dossiers.

Quantum stack dependency graph A quantum-company stack map with two correlated dependency points. Quantum stack dependency graph applicationssoftware + compilationcontrol + deviceinfrastructure access volumerisk
Figure 73.1. Company categories occupy separate layers. Backend access and production volume connect otherwise different portfolio nodes.

Classify a portfolio without ranking unlike firms

Prompt. Map six companies or projects across at least three stack layers using only dated primary evidence.

Deliverable. Node-edge table with category, buyer, revenue evidence type, upstream dependency, interface, and source date.

Pass condition. No cross-category leaderboard is produced, every edge is sourced, and the concentration analysis identifies at least one shared bottleneck.

Model answer: three clusters, two shared bottlenecks

Format. Anonymized six-node map with hardware access and procurement as separate concentration points.

The reference answer produces three peer clusters and flags two shared dependencies. It refuses to rank the control supplier against an application firm. The application cluster can be compared on buyer, outcome evidence, and backend substitution; the hardware cluster can be compared on workload-normalized evidence. The graph remains dated and is refreshed when an interface, supplier, or buyer contract changes.

Verification. The graph check finds the documented shared dependencies and confirms all comparison groups have matching buyer and layer fields.

Companion work

Artifacts for this chapter

These entries resolve to checked-in local source. Commands are reproduced exactly from the chapter manifest, and source-embedded fixtures are exported as direct downloads.

  1. Reproduce or test

    python3 tools/validate_briefs.py --briefs data/editorial_briefs_64_87.json --from 64 --through 87 --check-rewritten-sources --execute-artifacts

Provenance

Sources and review

  1. U.S. Government Accountability Office. Quantum Computing and Communications: Status and Prospects. GAO. 2021government technology assessment
  2. OpenQASM Technical Steering Committee. OpenQASM 3 specification. Linux Foundation Joint Development Foundation. 2026official technical specification
  3. IBM Quantum. Qiskit documentation. IBM. 2026official documentation
  4. Google Quantum AI. Cirq documentation. Google. 2026official documentation
  5. Association for Computing Machinery. Artifact Review and Badging. ACM Publications. 2026official reproducibility policy
  6. Easwar Magesan, J. M. Gambetta, and Joseph Emerson. Scalable and robust randomized benchmarking of quantum processes. Physical Review Letters. 2011primary paper
  7. Lieven M. K. Vandersypen et al.. A look at the full stack. Nature Reviews Physics. 2021peer-reviewed perspective
  8. Timothy Proctor et al.. Benchmarking quantum computers. Nature Reviews Physics. 2025peer-reviewed perspective

The load-bearing claims in the chapter are mapped inline to this registered source set. A citation supports only the bounded claim beside it.

Cite this chapter