Previous BLOG items described a small experiment aimed at demonstrating how eValid can test the performance delivered to a particular smartphone. You can read about that experiment in detail in our Mobile Agent Loading Experiment #1 writeup.
To see if there had been any changes as the result of our experiment we reran the same test on the same application approximately 3 weeks after the first run. We're pleased report that improvements to that application's server stack has apparently resulted in a a 50% improvement in performance and capacity. You can see the relative improvement in the two charts shown in this Comparative Improvement of Performance.