Time to get this show on the road, but first a recap on the tested speeds:
- 1600MHz C9-9-9-27 1T command rate
- 2400MHz C10-12-12-31 2T command rate ( XMP profile )
- 2400MHz C9-11-12-27 1T command rate ( optimised settings )
As usual starting off with SuperPi 32M. Over 12 secs faster computation for the 2400C10 XMP speeds over the 1600C9 simulated kit. If we tighten things up we shave off another 7 seconds.In Wprime 1024, it all remains within the margin of error.
In AIDA64 we can spot the huge differences in bandwidth once we up the speed from 1600MHz to 2400MHz. With the tighter timing set we achieve over 1300MB/s gain in the COPY test and over 2000MB/s increase for the READ test.
With the CPU Queen test in AIDA64 we see a minimalistic increase. However the PhotoWorXX test, which performs different common tasks during digital photo processing, sees a good boost coming from 1600MHz and scaling with less impressive with the tighter timings. Similar to the CPU Queen test we see no real benefits for fast RAM with Cinebench R10 64bit. It seems clear that it depends purely on the programs and operating system you are using if you will notice a performance gain yes or no.
In the encoding test, higher RAM speed reflects in more rendered frames per second, five frames might not mean much to you, but for those that work with long clips it can mean a world of difference. Nice scaling coming from 1600C9 and on top a solid 2.7 FPS gain in Pass 1 with the "tweaked" setup. That ultra fast rams hardly are beneficial for gaming purposes is nicely reflected in the 3DMark06 and game tests. 3DMark11, being heavy bandwidth dependant, receives a nice improvement in the Physx test. However the total score's improvement is far from massive.