Hey does anyone know where I can find some benchmarks for latency and speed? I want to know if getting 800 will really improve non-overclocked performance over 667.
Thanks
Thanks
There aren't any such results in the article I linked to. In fairness, the wording of the graphs is not always very good; for some of them the best score is the lowest (bar furthest to the left), so you might want to re-read the graphs.Okay some of those benchmarks didn't make any sense... EX: 667 5-5-5-15 beating 800 4-4-4-12,...
I don't know what CPU and MB you have, but my general answer would be "yes"; it doesn't hurt to experiment.Should I overclock my memory to 800 and tighten timings?
Like I said, the wording on the graph is very confusing, but it shows the DDR2-667 CL5 being the 2nd *worst*, better only than DDR2-400 CL5. The DDR2-800 CL4 is the *best* of all of them. Shorter times (faster completion of benchmark) are better; shorter times are to the left of the graph, longer times are to the right.No it was right to left, on the second page w/super pi.
It shouldn't. Most of the heat is carried away through the metal traces to the MB, anyway (the spreader doesn't do much if the chips are in [non-heat-conductive] plastic packaging, which almost all are).You think the memory will overheat if I overclock it?