1: \begin{abstract}

2: In this paper we present the solver DuQuad  specialized for solving

3: general convex quadratic problems arising in many engineering

4: applications. When it is difficult to project on the primal feasible

5: set, we use the (augmented) Lagrangian relaxation to handle the complicated

6: constraints and then, we apply dual first order algorithms based on

7: inexact dual gradient information for solving the corresponding dual

8: problem. The iteration complexity analysis is based on two types of

9: approximate primal solutions:  the primal last iterate and an

10: average of primal iterates. We provide computational complexity

11: estimates on the primal suboptimality and feasibility violation of

12: the generated approximate primal solutions. Then, these algorithms

13: are implemented in the programming language C in DuQuad, and

14: optimized for low iteration complexity and low memory footprint.

15: DuQuad has a dynamic Matlab interface which make the process of

16: testing, comparing, and analyzing the algorithms simple. The

17: algorithms are implemented using only basic arithmetic and logical

18: operations and are suitable to run on low cost hardware. It is shown

19: that if an approximate solution is sufficient for a given

20: application, there exists problems where some of the implemented

21: algorithms obtain the solution faster than state-of-the-art

22: commercial solvers.

23: 

24: \keywords{Convex quadratic programming \and (augmented) dual relaxation \and

25: first  order algorithms, \and rate of convergence \and arithmetic

26: complexity.}

27: \end{abstract}

28: