¿QUÉ ES LA COMPUTACIÓN CUÁNTICA?
Otra introducción a la computación cuántica...
2. What’s Quantum Computing?
2.1 Más que lanzar monedas
El ejemplo de las monedas que se lanzan al aire puede ser útil para explicar cómo se comportan los efectos cuánticos, pero sería poco práctico utilizar monedas como método de procesamiento cuántico de la información: la computación cuántica. Hay algunas razones obvias para ello:
- Lanzar monedas no es muy rápido. Los ordenadores convencionales (CPU) tienen una velocidad de reloj del orden de los GHz o miles de millones de ciclos de reloj completos por segundo. Necesitaríamos un número astronómico de monedas para intentar replicar eso.
- Flipping coins can only really be used to generate random numbers. Beyond the conventional deterministic computing operations we can already do, the only real advantage from quantum theory would be the ‘random’ nature of the coin flips.
- Juntar las monedas es aún más lento que darles la vuelta. Por no hablar de la parte en la que se barajan de forma pseudoaleatoria.
En otras palabras, un ordenador cuántico práctico tendría que ser, como cabría esperar, un sistema cuántico que pudiera controlarse con eficacia y eficiencia para realizar cálculos útiles.
2.2 ¿Qué hacen los computadores?
Antes de añadir especias cuánticas a nuestros ordenadores, sería prudente describir primero qué es lo que hacen nuestros ordenadores clásicos.
Computers work by executing logical operations (think addition/subtraction/multiplication/…) that take the machine from a starting point to a finishing point. The starting point that is fed into the computer is conventially known as the input and what the computer ends up with is usually referred to as the output. An *algorithm* is a set of instructions that allow the computer to take the input and generate an output. For example, an algorithm for doubling a number could take an input of 3, multiply it by 2 and return an output of 6.
Going back to the coins here, this is a bit like laying them out in a row and then proceeding to shuffle and flip the coins according to a script that allows something to be done. For example, you could add 2 numbers expressed in binary by following a simple series of rules that allows for addition. This is not too disimilar to using an abbacus as a calculator. Since the advent of the digital computer in the 1940’s, these operations have been carried out by increasingly advanced machines to greater and greater success.
Esta descripción de la computación es lamentablemente inexacta para los estándares modernos, pero basta para esbozar los fundamentos básicos y es suficiente para motivar el uso de sistemas cuánticos.
2.3 Algoritmos cuánticos
Los ordenadores convencionales son realmente buenos en lo que hacen. El campo de la computación de alto rendimiento ha tenido muchas décadas para madurar. Según Top500, un índice de seguimiento de los superordenadores más potentes, Fugaku, el superordenador más potente, es una potencia de 1.000 millones de dólares. Consume 29.899 kilovatios para alimentar 7.630.848 núcleos. En comparación, un portátil de gama media puede tener 6 núcleos y consumir unos 30 vatios en total.
For both machines, they execute algorithms that take an input and generate outputs in a similar manner. Whilst they are both excellent at solving a great many problems, there are two things they can’t handle so well: superposition & entanglement.
Starting with a simple coin toss there are 2 possible outcomes. If we add a second coin there are 4 possible outcomes (HH, HT, TH, TT). As more and more coins are added to the tosser, the number of possible outcomes, $N$, grows as $latex N = 2^n$. If we had 300 coins, that would mean a single coin toss has more possible outcomes than we estimate the total number of atoms in the universe. That’s roughly 2 followed by 90 0’s different combinations of H & T. Now imagine tossing these coins 300 times…
Un ordenador cuántico podría resolver este problema fácilmente. En lugar de necesitar un universo de átomos, bastaría con 300 qubits *de calidad suficientemente alta* para ejecutar un algoritmo cuántico que pudiera simular estas monedas (incluyendo cualquier pegado extravagante de las mismas) con una precisión arbitraria.
Quantum algorithms are similar to the input and output of classical algorithms but with the addition of superposition and entanglement in the middle. If an algorithm doesn’t feature these two concepts, a classical solution will outperform any quantum computer any time.
2.4 Ventaja Quantum
If our conventional computers perform so many tasks better, there must be some motivation for quantum computing being such an exciting area of research. Aside from the elegance of manipulating quantum states, quantum computing is expected to be more useful than a glorified coin tossing simulator.
Despite the limitations in state of the art quantum computing machines, it is known of a few valuable applications where quantum algorithms *theoretically outperform* the best possible classical solution. In these specific applications it is said there is a proven *quantum advantage* as using the quantum computer derives some benefit to the user for that application.
Whether or not it is worth using a quantum computer for a specific task is a very complicated question. It is not as simple as finding out if a quantum computer is faster but also whether the output of the quantum computer is better than what could have been done with conventional algorithms on a powerful classical computer. As of the time of writing, there has been no practical demonstration of a problem with real world (outside experimental physics) problem where using a quantum computer was better than using the tried and tested classical solution. As quantum computers become increasingly powerful it is hoped that this threshold will be crossed in the next few years…
Autor
Thomas Clarke – Physicist and MSc in Quantum Computing
ARTÍCULOS CUÁNTICOS
Artículos que pueden ser de tu interés
