With so many benchmarks to consider, grouping them together will help determine average performance differences for each application or suite. Let's start with games.
Warhammer was one of two games where the low-cost system failed to deliver the minimum required performance. The mid-priced system was twice as powerful, while the high-end system was almost three times as fast. Even at stock speeds, the smallest difference between the low-cost and high-end system was 75%.
With a lead of up to 169% over the slowest configuration, the high-end overclocked system starts to look like a decent value in some applications.
Synthetics are intended to represent a broader range of operating conditions than a typical benchmark set, but we occasionally have our doubts about their value. Sandra shows a spread of 167%, while PC Mark 2005 shows only a 47% difference between the fastest and slowest systems.
Grouping the results further by application type will help define how each system fits a particular market.
Spending large quantities of money pays off best in games, but isn't that what everyone expected? The high-end system also leads the slowest configuration in applications by 54% at stock speed and 110% when overclocked. The overclocked mid-priced system's application performance actually passes the high-end system's base-speed performance.