Last time when we looked at the KT4 Ultra, going from DDR333 to DDR400 settings actually reduced our benchmark results. Our speculation at the time was that our DDR400 RAM was running at CL2.5, while results from our 333 RAM were done at CL2.
Well now Corsair has supplied us with their XMS3500 RAM that runs at 400Mhz/CL2. So weve performed some benchmarks to see exactly what happens.
Looking at these new and updated results reasserts our previous findings: DDR400 simply offers no performance advantage over DDR333. And now we can even refute our previous question of whether a lower CAS latency would have allowed DDR400 to compete against DDR333.
One important note for those of you following DDR400: notice how Soyos DDR400 scores did indeed rise when we moved from CL2.5 to CL2. And though not shown here, these scores reflect an almost 15% or more increase from scores that we recorded when using the original BIOS that had shipped with the Soyo. This is significant for 2 reasons:
1) DDR400 performance can improve in the future, with BIOS and possibly chipset improvements.
2) CAS latency does play a role, even if it cant currently boost scores enough to rattle DDR333 results
In spite of these revelations, DDR400 simply is not fit for use with the AMD platform at the moment, and Id advise that everyone stick with DDR333.
forums
send article 
hardware newsletter 
article comments (2)
Digg this!
del.icio.us