1: \begin{abstract}

2: We survey old and new results about optimal algorithms for summation of 

3: finite sequences and for integration of functions from 

4: H\"older or Sobolev spaces.  First we discuss optimal deterministic and 

5: randomized algorithms. Then we add a new aspect, which has not been 

6: covered before on 

7: conferences about (quasi-) Monte Carlo methods: 

8: quantum computation. We give a short introduction

9: into this setting and

10: present recent results of the authors on optimal quantum algorithms

11: for summation and integration. We discuss comparisons between the three settings. 

12: The most interesting case for Monte Carlo and

13: quantum integration is that of moderate smoothness $k$ and large

14: dimension $d$

15: which, in fact, occurs in a number of important applied problems. In

16: that case the

17: deterministic exponent is negligible, so the

18: $n^{-1/2}$

19: Monte Carlo and the $n^{-1}$ quantum speedup essentially constitute the

20: entire convergence rate. We observe that 

21: \begin{itemize} 

22: \item 

23: there is an exponential speed-up of quantum algorithms over deterministic (classical)

24: algorithms, if $k/d$ tends to zero;

25: \item

26: there is a (roughly) quadratic speed-up of quantum algorithms over randomized

27: classical algorithms, if $k/d$ is small. 

28: \end{itemize} 

29: \end{abstract}

30: