quantum computing architecture
A normal computer stores information in bits, each definitely 0 or 1. A quantum computer uses quantum bits, or qubits, which exploit two strange rules of quantum physics. Superposition lets a qubit be in a blend of 0 and 1 at once until it is measured, and entanglement lets qubits become correlated so that they must be described together, not separately. With n qubits the machine can, in a precise mathematical sense, manipulate a state space of 2^n possibilities at the same time. This sounds like magic, and the magic is real but narrow — and the first job of an honest description is to fence off what it is not.
Here is the careful boundary, because the hype is relentless. A quantum computer is not a faster classical computer — it will not make your spreadsheet, your web browser, or ordinary code run quicker, and for most everyday tasks a classical machine is better. It is not a magic box that 'tries all answers at once' and reads off the best: you cannot directly observe the 2^n amplitudes; measuring a qubit collapses it to a single 0 or 1, and the art of a quantum algorithm is arranging interference so that the right answers reinforce and wrong ones cancel before you measure. And it is not a general solver for NP-complete problems — there is no known quantum algorithm that solves them efficiently in general. The proven speedups are specific: factoring large numbers (Shor's algorithm, which threatens some current cryptography) and a quadratic — merely square-root, not exponential — speedup for unstructured search (Grover's algorithm), plus genuine promise for simulating quantum chemistry and physics.
As architecture, the engineering is brutally hard, and this is why it is a frontier rather than a product. Qubits are exquisitely fragile: the tiniest stray heat, vibration, or electromagnetic noise disturbs them, an effect called decoherence, so they must be isolated and often chilled to near absolute zero. They make errors constantly, so a working machine needs quantum error correction, which may require hundreds or thousands of physical qubits to build one reliable logical qubit. Today's devices are small and noisy — the so-called NISQ era (noisy intermediate-scale quantum) — useful for research but not yet for large practical problems. A realistic view: quantum computing is a profound long-term bet that, if it matures, will transform a few specific domains, while leaving the vast bulk of computing exactly where it is.
Grover's search is the textbook example of an honest quantum speedup — and its limits. Searching an unstructured list of N items classically takes about N steps; Grover's algorithm takes about the square root of N. For a trillion items that is a million steps instead of a trillion — a real, useful gain, but a quadratic speedup, not the exponential 'instant answer' the headlines imply. And it still requires a large, error-corrected quantum machine that does not yet exist at scale.
Even quantum's clear wins are bounded — Grover gives a square-root, not exponential, speedup, and needs hardware we do not yet have.
A quantum computer is not a faster classical computer, does not try all answers at once and read off the best, and is not a general solver for NP-complete problems. Its proven speedups are narrow (factoring, quadratic search, quantum simulation), and today's machines are small and noisy (the NISQ era).