The number of parallel systems using accelerators is growing up. The technology is now mature enough to allow sustained petaflop/s. However, reaching this performance scale requires efficient scheduling algorithms to manage the heterogeneous computing resources. We present a new approach for scheduling independent tasks on multiple CPUs and multiple GPUs. The tasks are assumed to be parallelizable on CPUs using the moldable model: the final number of cores allotted to a task can be decided and set by the scheduler. More precisely, we design an algorithm aiming at minimizing the makespan---the maximum completion time of all tasks---for this scheduling problem. The proposed algorithm combines a dual approximation scheme with a fast integer linear program (ILP). It determines both the partitioning of the tasks, ie whether a task should be mapped to CPUs or a GPU, and the number of CPUs allotted to a moldable task if mapped to the CPUs. A worst case analysis shows that the algorithm has an approximation ratio of $\frac{3}{2} + \epsilon$. However, since the complexity of the ILP-based algorithm could be non-polynomial, we also present a proved polynomial-time algorithm with an approximation ratio of $2+\epsilon$. We complement the theoretical analysis of our two novel algorithms with an experimental study. In these experiments, we compare our algorithms to a modified version of the classical \heft algorithm, adapted to handle moldable tasks. The experimental results show that our algorithm with the $\frac{3}{2} + \epsilon$ approximation ratio produces significantly shorter schedules than the modified \heft for most of the instances. In addition, the experiments provide evidence that this ILP-based algorithm is also practically able to solve larger problem instances in a reasonable amount of time.