Quantum computers: how do they work?

This art­icle is part of our spe­cial issue « Quantum: the second revolu­tion unfolds ». Read it here

Uni­ver­sit­ies and labor­at­or­ies around the world, as well as the big tech com­pan­ies, Google, IBM, Intel and Microsoft, are study­ing and devel­op­ing quantum com­puters at break­neck speed. The stakes are high. But what exactly is this technology?

A quantum com­puter is not really a ‘com­puter’ as such, but rather a super-cal­cu­lat­or, cap­able of run­ning spe­cif­ic power­ful quantum algorithms much faster than an ordin­ary pro­cessor. It does this using the prin­ciples of quantum mech­an­ics, which are respons­ible for the beha­viour of ele­ment­ary particles such as photons, elec­trons and atoms, but also of lar­ger sys­tems such as super­con­duct­ing circuits.

Such sys­tems allow for the imple­ment­a­tion of quantum bits (‘qubits’): two-state quantum sys­tems that rep­res­ent the fun­da­ment­al com­pu­ta­tion­al bricks of quantum information.

Unlike con­ven­tion­al com­puters, which encode inform­a­tion in bin­ary form, qubits are not lim­ited to ‘0’ and ‘1’ but can be in any com­bin­a­tion (or ‘super­pos­i­tion’) of the two. This abil­ity, com­bined with the fact that N qubits can be com­bined or ‘entangled’ to rep­res­ent 2^N states sim­ul­tan­eously, allows cal­cu­la­tions to be per­formed in par­al­lel on a massive scale. A quantum com­puter could there­fore, in prin­ciple, out­per­form a clas­sic­al com­puter for some import­ant tasks, such as sort­ing large unsor­ted lists or for ‘prime fac­tor­isa­tion’. The lat­ter forms the basis of most encryp­tion algorithms in use today – in par­tic­u­lar for bank­ing operations.

Quantum com­puters look more like large cans, sus­pen­ded from the ceil­ing, cooled to near abso­lute zero with hun­dreds of cables hanging from them.

Quantum com­puters don’t look any­thing like their clas­sic­al coun­ter­parts either. Cur­rent mod­els look more like large cans that are sus­pen­ded from the ceil­ing, cooled to near abso­lute zero (-273.14°C), with hun­dreds of cables hanging from them.

Any­one wish­ing to build a quantum com­puter today must first over­come a big prob­lem: the fact that qubits are extremely fra­gile. Almost any inter­ac­tion with extern­al ‘noise’ in their envir­on­ment can cause them to col­lapse like a soufflé and lose their quantum nature in a destruct­ive pro­cess known as deco­her­ence. If this hap­pens before an algorithm has fin­ished run­ning, the res­ult is a com­plete mess (and not the res­ult of a cal­cu­la­tion) because any inform­a­tion stored in the qubit is lost (ima­gine a com­puter that has to restart every second). And, the more qubits a quantum com­puter has, the more dif­fi­cult it is to keep them coher­ent.  So much so that even the most advanced quantum pro­cessors today struggle to sur­pass 60 phys­ic­al qubits. A real device would require sev­er­al thou­sands, however.

Because of this prob­lem, a quantum com­puter must be well isol­ated from its envir­on­ment. This requires very strin­gent con­di­tions: simple and very cold sys­tems, pro­tec­ted from the out­side world. Such extreme con­fine­ment cre­ates a para­dox­ic­al situ­ation, how­ever, because the more isol­ated the com­puter is, the more dif­fi­cult it is for us to actu­ally com­mu­nic­ate with it (to access the res­ults of its cal­cu­la­tions) and con­trol what it does.

In recent years, qubits have been made from a num­ber of isol­ated sys­tems that remain coher­ent while algorithms are being run. These include trapped ions (atoms with an elec­tron removed or added), ultra-cold ‘Rydberg’ atoms, and photons (light particles).

A trapped-ion based com­puter stores inform­a­tion in the energy levels of indi­vidu­al ions. This is how qubits are formed in this sys­tem. Inform­a­tion is shared between these qubits, and laser pulses are then used to manip­u­late their states and cre­ate entan­gle­ment between them. This tech­no­logy is quite advanced and research­ers have recently suc­ceeded in cre­at­ing a fully entangled ‘24-qubit-GHZ’ (GHZ for ‘Green­ber­ger-Horne-Zeilinger’) state using cal­ci­um ions¹.  

Oth­er sys­tems are based on sol­id-state mater­i­als that might be integ­rated into tra­di­tion­al elec­tron­ic devices. These micron-sized struc­tures are small on every­day scales, but large com­pared with atoms and can behave like quantum particles, such as elec­trons or atoms. Such ‘arti­fi­cial quantum objects’ include ‘quantum dots’ (tiny pieces of semi­con­duct­ing mater­i­al), super­con­duct­ing cir­cuits, and dia­monds con­tain­ing spe­cial type of defects called ‘nitro­gen vacan­cies’. A quantum com­puter based on super­con­duct­ing qubits, for example, is cooled down to mil­likelvin tem­per­at­ures (which is colder than inter­stel­lar space) and is con­trolled using microwaves².

Research­ers are also try­ing to work out which of these sys­tems would make the best qubits. An import­ant para­met­er is, of course, a qubit’s res­ist­ance to deco­her­ence, which can be assessed in terms of the ‘fidel­ity’ of a quantum oper­a­tion. While the fidel­ity does not have to be a per­fect 100%, any­thing lower will even­tu­ally lead to errors after mul­tiple oper­a­tions; most of today’s quantum com­puters are very sus­cept­ible to errors.

Although ‘quantum error cor­rec­tion’ pro­to­cols can mit­ig­ate deco­her­ence, these are costly from a hard­ware point of view and any prac­tic­al sys­tem must have a suf­fi­ciently high fidel­ity to begin with. Research­ers are mak­ing pro­gress in this area, how­ever, and recent work has shown that a two-qubit gate can be made from two sil­ic­on quantum dots³. This gate can achieve 98% fidel­ity for the CROT oper­a­tion, which an essen­tial com­pon­ent of a quantum computer.

The idea of a quantum com­puter was first put for­ward by the late phys­i­cist and Nobel laur­eate Richard Feyn­man in the 1980s to sim­u­late the com­plic­ated equa­tions of quantum mech­an­ics, which take too long to solve on a clas­sic­al com­puter. Today, applic­a­tion areas are much more var­ied: cryp­to­graphy; sim­u­lat­ing the prop­er­ties of mater­i­als (with a view to improv­ing them); solv­ing dif­fer­en­tial equa­tions extremely quickly; and optim­iz­ing machine learn­ing. Pro­gress has been impress­ive and research­ers have gone from being able to entangle just three qubits to more than 50 qubits in recent years, with error rates of 1 in 1000⁴.

Although it is dif­fi­cult to pre­dict what the future holds, a fully func­tion­al, com­mer­cially viable quantum com­puter is unlikely to see the day of light any­time soon. The same goes for any kind of ‘per­son­al quantum com­puter’. A quantum device will more likely be used for fun­da­ment­al research, R&D or gov­ern­ment and mil­it­ary pur­poses to begin with.

Today, research­ers in this field are not only advan­cing hard­ware (i.e. the machines them­selves), but are also devel­op­ing innov­at­ive soft­ware with new types of algorithms espe­cially adap­ted to quantum computing.

While quantum com­put­ing relies on the prin­ciples of fun­da­ment­al phys­ics, it is a tre­mend­ous oppor­tun­ity for sci­ent­ists from many fields – includ­ing com­puter sci­ence, math­em­at­ics, mater­i­als sci­ence and engin­eer­ing – to work togeth­er. The road to a real-world quantum com­puter will cer­tainly be long, but there will be many excit­ing dis­cov­er­ies along the way. 

Interview by Isabelle Dumé