Quantum Speedup
When a quantum algorithm solves a problem asymptotically faster than the best known classical algorithm.
Quantum speedup refers to a quantum algorithm outperforming classical algorithms, either provably or based on current knowledge.
Types of Speedup
Proven Speedup
Quantum is provably faster (information-theoretic lower bound for classical):
| Problem | Classical | Quantum | Speedup |
|---|---|---|---|
| Unstructured search | Quadratic | ||
| Parity on OR tree | Quadratic |
Superpolynomial (Believed)
Based on computational complexity assumptions:
| Problem | Classical (best known) | Quantum |
|---|---|---|
| Factoring | ||
| Discrete log | ||
| Certain simulations | Exponential | Polynomial |
Polynomial Speedup
Faster by a polynomial factor:
| Problem | Classical | Quantum |
|---|---|---|
| Search | ||
| Collision finding |
The Speedup Landscape
Quantum Speedups
│
┌─────────┴─────────┐
│ │
Exponential Polynomial
│ │
Factoring Grover search
Simulation Many others
Period finding
Important Caveats
Input/Output
Many quantum speedups assume quantum input or only require quantum output:
- HHL: Exponential speedup… if you can prepare input and only need expectation values
- Speedup may vanish with classical I/O overhead
Constant Factors
Quantum algorithms often have large constant factors:
- Gate overhead
- Error correction costs
- Classical simulation may win for small problems
Problem Structure
Speedups exploit specific problem structure:
- Shor’s uses periodicity
- Grover’s uses function structure
- Not all problems have exploitable structure
No Speedup
Some problems have no quantum speedup:
- Sorting: Still
- General NP-complete: No known speedup
- PSPACE-complete: Quantum in PSPACE, so no exponential speedup
Measuring Real Speedup
For practical comparisons:
- Wall-clock time on real problems
- Include all overheads (error correction, I/O)
- Compare to best classical hardware and algorithms
Current status: Still demonstrating quantum advantage on artificial problems.
See also: Quantum Advantage, Grover’s Algorithm, Shor’s Algorithm