Many of us have been wondering about how to tweak SuperPI 32M, trying to figure out how you can be faster clock-per-clock and what memory subtimings we have to alter in order to speed up the benchmarking process. Next to the everlasting quest to squeeze out every possible MHz, there’s the back-to-back cas delay which everyone has been using so far. Next to this little subtiming, there’s the ‘Round Trip Latency’, or: RTL, which seems to have even more effect.
In this article posted on AnandTech; Raja, who is known from BenchTek, OCX and XS, gives a little bit more background to this particular memory timing and explains why tuning it can be beneficial for your benchmark session. I recommend any overclocking enthousiast to have a look at the article! Needless to say, all credits go out to Raja.
Link to the article: AnandTech