The Via Apollo Pro 133A chipset has a very nifty feature. It lets you run the memory at a different speed than that of the FSB. This is a boon for overclockers as well as those looking to put their old PC100 SDRAM to good use. But how much performance penalty can you expect by running the memory at a slower speed?

There is very little difference in CPU performance as this benchmark isolates the CPU from other subsystems when benchmarking. Similar results are attained when PC133 SDRAM is run at 100MHz.





There is a drop in performance as I expected there would be. But, the performance difference between PC133 and PC100 SDRAM is hard to tell during everyday computing. Also as expected, there is a similar drop in performance when PC133 SDRAM stick is run at 100MHz instead of 133MHz. That means that there is no point in buying a PC133 SDRAM if you are planning to run your memory at 100MHz.





This benchmark result made me stare at the screen scratching my head for an hour. Why is running the memory at a slower speed affecting hard drive performance? I thought there was something wrong with the benchmark so I repeated them several times. But I kept getting the same results. The only explaination I can think of is that when the memory is run at a slower speed, the FSB runs slower as well to some degree. In any case, as long as the memory is run at the same speed as the FSB, there shouldn’t be any slow downs in drive performance.

Pages: 1 2 3 4