1: \begin{abstract}

2: Gradient-type iterative methods for solving Hermitian eigenvalue problems

3: can be accelerated by using preconditioning and deflation techniques. 

4: A preconditioned steepest descent iteration with implicit deflation (PSD-id)

5: is one of such methods. 

6: The convergence behavior of the PSD-id is recently investigated 

7: based on the pioneering work of Samokish on the preconditioned 

8: steepest descent method (PSD).  The resulting non-asymptotic estimates indicate 

9: a superlinear convergence of the PSD-id under strong assumptions on the 

10: initial guess. The present paper utilizes an alternative convergence 

11: analysis of the PSD by Neymeyr under much weaker assumptions. 

12: We embed Neymeyr's approach into the analysis of the PSD-id using a 

13: restricted formulation of the PSD-id. More importantly, we extend the 

14: new convergence analysis of the PSD-id to a practically preferred 

15: block version of the PSD-id, or BPSD-id, and show the cluster robustness 

16: of the BPSD-id. Numerical examples are provided to validate the 

17: theoretical estimates. 

18: \end{abstract}

19: