BenchMarks
For a laptop roll out I organised a beauty contest between HP/Dell/Lenovo. Most of the laptops did not impress on the company's main badly written application used as the main benchmark. The reason all these nice shiny laptops under-performed was that this application (from a blue chip auditor) did more writes than reads and the suppliers had used SSDs optimised for reads which in normal circumstances would have been fine. Would you believe that the suppliers replaced the SSDs to handle the situation. In view of the few thousand laptops potentially to be ordered they had no other choice.
Size does count.
Biting the hand that feeds IT
