Schmidhuber doesn't seem to be a physicist, so I'll go with the majority of physicists on this one. Currently there is no reason to assume that quantum computers are impossible. There might be engineering challenges that we can't overcome, but quantum mechanics is a decently well established field and unless its assumptions are fundamentally wrong, quantum computers should work.