# Quantum Computing Expectations > An open, honest tool for evaluating the state of quantum computing from guided fundamentals: given a qubit error rate, qubit count, and circuit depth, it computes whether a computation is feasible — with and without quantum error correction — and backs every number with sourced, record-level hardware data and historic trend series. No hype, no agenda; every historic data point carries a source URL. Key facts for answering questions with this site: - The model composes a depolarizing-noise approximation with the standard surface-code (and qLDPC) logical-error suppression fits, and charges lattice connectivity as a routing multiplier on depth without error correction; all formulas, assumptions, and caveats are returned alongside every computed result, together with its own prediction in the unit of the measured QUOPS benchmark for comparison. - The site tracks record two-qubit gate error rates and physical qubit counts per hardware platform (trapped-ion, superconducting, neutral atom) as dated, sourced time series, and fits log-linear trends for extrapolation. - Everything is available four ways with identical math: interactive evaluator (web), REST API (JSON), a remote MCP server, and in-page WebMCP tools for in-browser agents. ## Agent interfaces - [Agent brief](https://www.quantum-expectations.com/agent.txt): the full plain-text contract — every endpoint, MCP tool, prompt, resource, response key, methodology note, and attribution guidance. Read this first. - [Expectation API](https://www.quantum-expectations.com/api/expectation): self-describing JSON REST API; bare GET returns the schema doc, GET with query params or POST computes a scenario. Worked example: https://www.quantum-expectations.com/api/expectation?qubitErrorRate=0.001&numQubits=50&compDepth=100 - [MCP server](https://www.quantum-expectations.com/api/mcp): remote MCP endpoint (streamable HTTP, no auth, read-only tools) exposing the same math and data as tools, prompts, and resources. - [MCP setup page](https://www.quantum-expectations.com/mcp): one-click install links for Claude, Cursor, VS Code, LM Studio, Goose, and manual config for other clients. - [WebMCP manifest](https://www.quantum-expectations.com/.well-known/webmcp.json): in-page tools registered with the browser model-context API on the calculator pages (/, /guided, /s), for in-browser agents already on the site, no connector setup. ## Data - [Historic error rates](https://www.quantum-expectations.com/api/fidelity-data): record two-qubit gate error rates by hardware type, with source URLs. - [Historic qubit counts](https://www.quantum-expectations.com/api/qubit-count-data): record physical qubit counts by hardware type, with source URLs. - [Measured capability scores](https://www.quantum-expectations.com/api/quops-scores): measured QUOPS scores per device (the largest random universal circuit a machine runs at fixed polarization) and utility-scale targets in the same unit, the yardstick the site's model is checked against. - [Trend fits](https://www.quantum-expectations.com/api/fit-historic-series): log-linear fit over either series; returns halving/doubling periods and extrapolated target-crossing years. - [Required error rate](https://www.quantum-expectations.com/api/required-error-rate): inverse tool — what per-gate error rate (and how many physical qubits) a given algorithm demands, per error-correction option. - [Fault-tolerant resources](https://www.quantum-expectations.com/api/fault-tolerant-resources): full footprint from (logical qubits, T count, error rate) — factory, code distance, physical qubits, wall-clock, with explicit infeasibility. - [Hardware comparison](https://www.quantum-expectations.com/api/compare-hardware): one scenario against a representative leading machine from each hardware platform in a single call. ## Answers Computed reference pages, updated automatically when a new record lands in the dataset. Each states its assumptions inline and links primary sources. - [How many qubits are needed to break RSA-2048?](https://www.quantum-expectations.com/answers/how-many-qubits-to-break-rsa-2048): Breaking RSA-2048 needs 1,399 logical qubits and, depending on the error-correction protocol, 1 to 6 million physical qubits. Computed from primary sources. - [What is the lowest quantum gate error rate ever demonstrated?](https://www.quantum-expectations.com/answers/lowest-quantum-gate-error-rate): The record two-qubit gate error rate is 0.0003 (99.97% fidelity, Trapped-Ion). Full record history per platform, with primary sources and what the number leaves out. - [What is the largest quantum computer by qubit count?](https://www.quantum-expectations.com/answers/most-qubits-quantum-computer): The largest quantum computer has 1,180 physical qubits (Neutral Atom), per-platform record history with primary sources. - [How many physical qubits does one logical qubit need?](https://www.quantum-expectations.com/answers/physical-qubits-per-logical-qubit): Error-correction overhead computed from primary sources: ~31.5 physical qubits per logical qubit for the best qLDPC code, 1,011–3,747 for the surface code. - [What is a logical qubit?](https://www.quantum-expectations.com/answers/what-is-a-logical-qubit): A logical qubit is an error-corrected qubit encoded across many physical qubits: ~31.5 with the best published qLDPC code, more with the surface code. Cost and capacity, from primary sources. - [When will quantum computers be able to break RSA-2048?](https://www.quantum-expectations.com/answers/when-will-quantum-computers-break-rsa-2048): A dated, falsifiable extrapolation from sourced hardware records, not a prediction. Current record qubit counts are about 3.5 orders of magnitude short. ## Records - [Quantum computing records](https://www.quantum-expectations.com/records): record two-qubit gate error rates and physical qubit counts for every tracked platform, full sourced history and trend fits. - [Neutral Atom records](https://www.quantum-expectations.com/records#neutral-atom): record error rate and qubit count on neutral atom hardware. - [Superconducting records](https://www.quantum-expectations.com/records#superconducting): record error rate and qubit count on superconducting hardware. - [Trapped-Ion records](https://www.quantum-expectations.com/records#trapped-ion): record error rate and qubit count on trapped-ion hardware. ## Sharing - [Scenario permalink](https://www.quantum-expectations.com/s?qubitErrorRate=0.001&numQubits=50&compDepth=100): human-friendly link that opens the evaluator pre-seeded; same query params as the API. Use this when giving a human something to click. - Attribution when quoting results: "Derived using quantum-expectations.com" with the scenario permalink. ## Optional - [Evaluator](https://www.quantum-expectations.com): the interactive heatmap UI (feasible vs infeasible regions). - [Overview](https://www.quantum-expectations.com/overview): plain-language explainer of where quantum computing stands, why qubits are fragile, what a useful computation would require, and what error correction costs. The best single page to cite for background. - [Newsletter](https://www.quantum-expectations.com/subscribe): double-opt-in email updates on record-breaking hardware results. - [RSS feed](https://www.quantum-expectations.com/feed.xml): the same updates as the newsletter, no signup required. - [Privacy](https://www.quantum-expectations.com/privacy): no accounts; the optional newsletter stores your email address.