EVGA X58 SLI LGA 1366 Motherboard Review

Memory Benchmarks

Everest Ultimate is the most useful tool for any and all bench markers or overclockers. With the ability to read most voltage, temperature, and fan sensors on almost every motherboard available, Everest provides the ability to customize the outputs in a number of forms for display on your desktop. In addition to this, the memory benchmarking provides a useful tool of measuring the changes to your memory sub-system when tweaking to measure the differences.

We start today's benchmarks with memory bandwidth and the perennial all-star in memory benchmarks is Everest Ultimate's included memory benchmark. Unfortunately, that crown seems to be loosely attached to the head of Everest with this i7 setup. Reports of inconsistent Everest bandwidth results with i7 are 100% confirmed. During benchmarking we were able to get the Read results to vary as much as 1000 MB/s. This is rather significant when looking at bandwidth of the same settings. In the results above, we were sure to take three of the lower results and didn't include any of the random 'high' results in the calculations. With that said, we also have a bit of an anomaly in the Copy results. As we can see, the stock settings outpace the overclocked settings which just shouldn't happen. Perhaps it is Everest causing the odd result, or, we are looking at some sort of change in latencies similar to that of strap changes or internal north bridge timings of past Intel chipsets. Needless to say, Everest has left more questions that it has answered and at some point we hope to get to the bottom of those questions some day, but for today, we will leave well enough alone and consider these results as a 'guideline' more than anything else.

Unlike the bandwidth numbers, the latency results were 100% repeatable and 100% impressive coming from even highly clocked dual channel DDR3 setups based on X48 or 790i chipsets. This new on-die i7 memory controller is already showing its strength in bandwidth and latency results, it will be interesting to see how this equates to real world performance considering our two setups really aren't light years apart as far as the memory sub-system is concerned...according to Everest Ultimate anyway.

SiSoft Sandra is a popular and well used benchmark in the industry but not really a friend of serious benchmarkers. The results SiSoft Sandra produces have been suspect at times basing the numbers it comes up with on system specs and not actual testing. The latest version of Sandra seems to be one of the few programs that appear to calculate memory bandwidth consistently so we decided to include it in today’s benchmarks. Like we have always said with SiSoft Sandra though, take these results for what they are and nothing more.

Since we really don't much about what these results represent or how SiSoft Sandra is getting these results, we will leave any conclusions that may come from these numbers to interpretation by the readers. We can say that results seemed to be quite consistent throughout testing and the results seemed to scale how we would expect. Unfortunately, that is part of the problem with SiSoft Sandra memory bandwidth numbers, they always scale exactly how they should.

ScienceMark is an almost ancient benchmark utility at this point in time and hasn't seen an update in a long time. It is, however, still a favorite for accurately calculating bandwidth on even the newest chipsets.

As odd as it may sound, the oldest and out of date memory bandwidth measurement tool we use here at HWC was the most consistent in benchmarking this new chipset. The bandwidth results provide similar numbers to what was expected and scaled appropriately, even with results we aren't looking at here today. You may say that we said the same thing about Sandra and that is what made those results a little too consistent, but the difference is that ScienceMark actually measures bandwidth and comes up with a result in MB/s. Sandra comes up with its numbers in an unknown method and produces results in GB/s...which we can only assume is some theoretical bandwidth number that the memory is capable of, because this setup certainly isn't pushing 30GB/s of bandwidth.

Needless to say, this section is a bit of a mess so let's leave well enough alone and move on to some more benchmarks that rely on complete system performance for results.