For all of you out there with theoretical and practical experience, what is the best hardware and server types for large-volume OCR conversion. What type of processors and other resources have you found most effective and why? Please give a short description of your OCR work and loads to illustrate your feedback. Thanks for your feedback.
