Adiabatic Evolution Algorithm Chukman So (19195004) Steven Lee (18951053) December 3, 2009 for CS C191 In this presentation we will talk about the quantum mechanical principle of a new quantum computation technique, based on the adiabatic approximation. Two examples of its application is introduced, and its equivalence to tradition unitary-based quantum computer is demonstrated. Adiabatic Evolution Algorithm • Formulation – For a Hamiltonian H – Characterised by a some parameter λ (think box size in particle-in-box problem) – Solve the eigensystem H ( ) n ( ) En ( ) n ( ) • Adiabatic approximation – If the parameter is t, n (t ) does not give the right evolution – e.g. Start from certain 3 (t 0) , at a later time, may not be 3 (t ) – But if “slow enough”, this is approximately true time evolution n (t 0) n (t ) • If n (t 0) is a ground state of the initial H (t 0) time evolution will yield the ground state of H (t ) Quantum Computation by Adiabatic Evolution, E. Farhi, J Goldstone et al, Los Alamos arXiv 0001106 Adiabatic Evolution Algorithm • A tool to obtain the fulfilling assignment to a clause – E.g. Solve A OR B – Start with some initial H (t 0) H i ground state (t 0) 0 where 0 is the ground state of Hi – We need a final H (t T ) H f with the fulfilling assignment as lowest energy state AB H 1 00 00 0 01 01 0 10 10 0 11 11 00 00 Violating assignment Energy = 1 Fulfilling assignment Energy = 0 (this may seem useless, but clauses like this can be added = AND’ed) – By slowly varying H(t) from t = 0 to T, the initial ground state can be evolved into a fulfilling final state – But how slow? Adiabatic Approximation • For a time-dependent Hamiltonian H(t) – Time-dependent Schrodinger equation (t ) t – For any given time, instantaneous eigenstates can be found H (t ) (t ) i H (t ) n (t ) En (t ) n (t ) – We can always expend (t ) instantaneously using these kets, treating t just as a parameter i t En ( t ') dt ' (t ) A (t ) (t ) e 0 n n n Introduced without lose of generality Introduction to Quantum Mechanics, Bransden & Ostlie 2006 Adiabatic Approximation – The exact functional form of An (t ) is governed by TDSE; to make use of it we need i En ( t ') dt ' (t ) e 0 A ' ( t ) ( t ) A ( t ) n (t ) An (t ) n (t ) n n n t t n i t En ( t ') dt ' H (t ) (t ) e 0 A (t ) E (t ) (t ) t n n i E ( t ) n n n – Putting into TDSE, two terms cancel, leaving i t En ( t ') dt ' m (t ) e 0 A 'n (t ) n (t ) m (t ) n i t A 'm (t )e 0 Em ( t ') dt ' 0 En (t ') dt ' A (t ) (t ) e n n n t i t En ( t ') dt ' 0 m (t ) e A ( t ) ( t ) n n t n i t ( En ( t ') Em ( t ')) dt ' 0 A 'm (t ) An (t )e m (t ) n (t ) t nm i t Need to find this: TISE Adiabatic Approximation n (t ) for m ≠ n, we differentiate TISE on both sides by t t H ( t ) ( t ) En (t ) n (t ) n t t E H m n H n m n n En n t t t t H m n Em m n En m n for m n t t t 1 H 1 H m n m n m n t En Em t mn t – To find m (t ) – Putting this back, we have A 'm (t ) An (t ) nm 1 mn i mn dt ' H m n e 0 t – So far everything is exact – no approximations t Adiabatic Approximation A 'm (t ) An (t ) nm i mn dt ' H m n e 0 t t 1 mn – Adiabatic Approximation • Assuming the initial wavefunction is a pure eigenstate i only one Ai (t 0) 1 , all other zero • Assuming (a priori) at later time, other amplitudes stay small i.e. Ai (t ) 1 for all time, all other 0 (justified later by looking at the evolution) – Then we can simplify: A ' f i (t ) 1 fi i fi dt ' H f i e 0 t t – Integrating with time: i fi ( t '') dt '' H (t ') Af i (t ) dt ' f i e 0 0 fi (t ') t ' t 1 t' Adiabatic Approximation i fi ( t '') dt '' H (t ') Af i (t ) dt ' f i e 0 0 fi (t ') t ' t' 1 t – Now we can try to justify our a priori assumption – a crude way to approximate the order of this integral: ignore time dependence Af i (t ) 1 f fi (t ) H (t ) i t t 0 dt ' e i fi ( t ) t ' ei fi (t )t 1 H (t ) i i (t ) fi (t ) f t fi 1 Pf (t ) Af i (t ) 2 H (t ) i f 2 4 fi (t ) t 1 H (t ) i f 2 4 fi (t ) t 2 H (t ) 2 4 f i fi (t ) t 2 2 4 1 cos( 2 e fi i fi ( t ) t (t )t ) 1 2 Adiabatic Approximation – i.e. For our a priori assumption to work, we require H (t ) Pf (t ) 2 4 f i fi (t ) t 4 H (t ) i 2 fi 4 (t ) f t 4 H ( ) i ( E f Ei ) 2 f 1 f , t 2 d dt H ( ) i ( E f Ei ) 2 f 2 1 Putting back fi E f Ei 1 where (t ) t T T is the total ramp time from i to f state T – For adiabatic approximation to work, T must be large enough / ramp slow enough Adiabatic Approximation – This measure is important • Determines how fast the computation can be performed • Since 1 T ( E f Ei )2 the smaller the gap is, the more likely a transition is • The 1st excited state dominates • T chosen wrt. smallest gap during evolution • If states cross & matrix element f H ( ) i non-zero → computation fail which makes choosing the initial H i important What is SAT? • Boolean satisfiability problem (SAT) C1 C2 CM – Clause: A disjunction of literals Ck x1 x2 – Literal: a variable or negation of variables • Basically a huge Boolean expression, which we try to find a valid set of values for the variables to make the given problem TRUE overall • Adiabatic approximation setup: – N-bit problem maps to n variables; use time evolution to solve for problem • SAT is NP-complete NP-complete • Nondeterministic polynomial time (NP) – Verifiable in polynomial time by deterministic Turing machine – Solvable in polynomial time by nondeterministic Turing machine • NP-complete is a class of problems having two properties: – Being NP – Problem (in class) can be solved quickly (polynomial time) → all NP problems can be solved quickly as well • Showing that a NPC problem reducing to a given NP problem is sufficient to show the problem is NPC • P != NP? So far most believe that is not the case, thus NP-complete problems are at best deterministically solvable in exponential time SAT quantum algorithm • Create a time-dependent Hamiltonian which is a linear ramp between the initial/starting Hamiltonian and final/problem Hamiltonian – Idea is to, given enough time T, to slowly evolve the initial ground state (easy to find) to final ground state (hard to find) t t H t 1 H i H T T H s 1 s H i sH f f • Note n-bit SAT problems mean that the Hamiltonian we are working with exist in a Hilbert space spanned by N = 2n basis vectors • Thus finding ground state of problem Hamiltonian in general requires exponential time • Adiabatic approximation efficiency all depends on T, which is related to gmin Quantum Computation by Adiabatic Evolution, E. Farhi, J Goldstone et al, Los Alamos arXiv 0001106 Initial Hamiltonian Hi • Set up an initial Hamiltonian whose ground state is easy to find k Hi 1 2 1 k k with x x 0 1 1 0 H i k xk x x xk x 1 1 xk 0 and 2 1 1 1 xk 0 2 1 • Noticing that 3-SAT is equivalent to SAT: H i ,C H B c H B c H B c i H i H i ,C C j k Initial Hamiltonian Hi • Ground state for HB is xk = 0 for all kth bit x1 0 x2 0 1 xn 0 n /2 2 z 1 z1 z2 z2 zn n H i d k H i k k 1 • Reason why we use a ground state in the x-axis instead of the z-axis is to prevent gmin from becoming zero, else adiabatic approximation fails zn Problem Hamiltonian Hf hC ziC , z jC , zkC • Energy function of clause C: 0 if the bits satifsy the clause, else 1 • Total energy can be defined as sum of individual HC’s • Hf can be defined as follows: H f ,C z1 z2 zn h z C iC , z jC , zkC z1 z2 zn H f H f ,C C • Ground state is solution to SAT problem • If no solution exists, will minimize number of violated clauses (lowest energy) 1-bit problem • Consider a problem with one 1-bit clause satisfied with 1 bit H i H i1 12 Hi 1 2 Hf 0 0 12 1 2 1 0 Hf 0 0 • Setup time-dependent ramped Hamiltonian H s 1 s H i sH f 1 1 s s 1 H s 2 s 1 1 s • Eigenvalues: 1 E 2 E s (1 s ) 0 2 1 1 2 s (1 s) E 2 1-bit problem E 1 2 s (1 s ) E1 E 2s 2 2s 1 1 1 2 s(1 s) 2 E0 g min 1 2 when s 12 1 1 2 s (1 s ) 2 Quantum Computation by Adiabatic Evolution, E. Farhi, J Goldstone et al, Los Alamos arXiv 0001106 Grover Problem • A quantum search problem – Locate a specific entry in unstructured database 0 – Using the following notation for states ... mz1 mz 2 mz 3 ... mzn A total of n bits, each a spin measure in z – Given a quantum oracle Hamiltonian 1 Hf 0 H f 1 0 0 if 0 if 0 Lowest energy state – To find 0 , we start with an initial Hamiltonian for which ground state is known n 0 1 1 if h Hi 2 i.e. Lowest state = Hadamard state if h 0 Hi 1 h h How fast is Adiabetic Quantum Computation?, W. van Dam, M. Mosca et al, Los Alamos arXiv 0206003 Grover Problem – Linear ramp between the two Hamiltonian: t t H (t ) 1 H i H f 1 s H i s H f T T (1 s)(1 h h ) s(1 0 0 ) 1 (1 s) h h s 0 0 – Using Adiabatic Approx. • Solve instantaneous eigenvalues • Find out two lowest states separation • Get the bound on ramping rate H ( ) i ( E f Ei ) 2 f T 1 E 2 where s t T Grover Problem • Solve instantaneous eigenvalues E H E (1 s) h h s 0 0 – dot both sides with h and 0 ( E 1) h (1 s ) h s h 0 0 ( E 1) 0 (1 s ) 0 h h s 0 ( E s ) h s h 0 0 ( E 1 s ) 0 (1 s ) 0 h h ( E s ) h s h 0 1 s 0 h h E 1 s ( E s )( E 1 s ) s (1 s ) 0 h 2 or h 0 corresponding to E=1 roots need to solve this Grover Problem – Where the dot product 0 h 2 is well defined: 0 1 0 h mz1 mz 2 ... mzn 1 2n /2 m 0 z1 1 2 m 0 z2 1 0 1 2 ... ... m 0 zn 0 1 2 1 mz1 must be either 0 or 1 1 2n /2 – Therefore the eigenvalues are given by ( E s)( E 1 s) s(1 s) 0 h E 2 E s (1 s ) s (1 s )2 n E 2 E s (1 s )(1 2 n ) 0 1 1 4 s(1 s)(1 2 n ) E or 1 2 2 as Grover Problem • Eigenvalue spectrum, from n=2 to 20 • Against s (or time) Energy E 1 n-2 degenerate states E 1 1 4s(1 s)(1 2 n ) E 2 1st excited state 1 4s(1 s)(1 2 n ) 1 1 4s(1 s)(1 2 n ) E 2 ground state s Grover Problem – i.e. t H ( ) i ( E f Ei )2 f 1 E 2 n where E 1 4s (1 s )(1 2 ) E E 1 4s (1 s )(1 2 n ) s Grover Problem – Allowing the ramping rate to adjust to the gap 1 s 0 E 2 ds 1 1 s 0 1 4s(1 s)(1 2 n ) ds 1 1/2 1 1 ds du s 0 u 1/2 1 1 1 1 s2 s 1 u2 4 4 4 2 1/2 1 2 1/2 1 du du u 0 1 1 2 u 0 2 u 2 u 4 1 n /2 2 tan 1 2 2 8 T t 1 2 1 u s 4 4 1 2 n 4 1 1 1 1 2 n 4 4 2n 1 4 4 1 2 n 2 – Compared with conventional search, Tconventional total bit combinations Tquantum 2n/2 i.e. quantum quadratic speed up 2n Grover Problem • Choice of initial Hamiltonian Hi is important – Bad choice changes gap dependence → longer ramp time – e.g if we choose n 1 H i 1 x(i ) i 1 2 – Eigenvalues calculation in quant-ph/0001106 Energy 1st excited state Tquantum 2n No quantum speed up ground state s Quantum Computation by Adiabatic Evolution, E. Farhi, J Goldstone et al, Los Alamos arXiv 0001106 Approximating adiabatic with unitaries • Discretize the interval 0 to T into M intervals – Unitary written as product of M factors d i U t , t0 H t U t , t 0 dt T U T , 0 0 where T / M U T , 0 U T , T U T , T 2 U , 0 • Note that we want to make the intervals small enough so that the Hamiltonian is near-constant in each discrete interval if H t1 H t2 1 M t1 , t2 l , l 1 then U l 1 , l e iH l Quantum Computation by Adiabatic Evolution, E. Farhi, J Goldstone et al, Los Alamos arXiv 0001106 Approximating adiabatic with unitaries • Then we substitute the Hamiltonian with the ramped Hamiltonian between the initial and final Hamiltonians H f Hi 1 • M is T times polynomial in n • Trotter formula for self-adjoint matrices: ( A B ) e lim e A/ n B / n n n e • Thus n in the above equation needs to be large enough to be used as sufficient approximation Approximating adiabatic with unitaries H l uHi vH f where u 1 l T , v l T • Then by using a large K, we can approximate using Trotter formula: if K then e M 1 Hi H f iH l e iuH i / K ivH f / K e 2 K Approximating adiabatic with unitaries • Thus the whole equation can be written in 2K terms, half being each of these terms: e iu H i K or e ivH f K • Hi is sum of n commuting 1-bit operators, so related unitary can be written as product of n 1-qubit unitary operators • Hf is sum of commuting operators (each for each clause), so related unitary can be written as product of unitary operators, each acting only on qubits related to clause • Thus total number of factors is T2 times polynomial in n – If T is polynomial as well, then number of factors is also polynomial Conclusion • We have talked about: – Physical principle of quantum adiabatic evolution algorithm – Its equivalence to traditional unitary quantum computation – Its application in two examples: a one-bit SAT problem, and Glover problem • Much like tradition QC – Adiabatic evolution leads to quantum speed up in specialised problems – “Smartness” is needed • picking unitary vs picking initial Hamiltonian – No general rule