Not As What It Seems

Before jumping into the full suite of benchmarks for both AMD and Intel, there are some performance concerns that need to be addressed. G.SKILL has a fairly big problem on hand. As it stands, this particular DDR5-8000 Trident Z5 Royal kit will not receive a recommendation due to poor benchmark results using its 8000 MT/s XMP profile. G.SKILL has been notified as well. Hopefully a revision comes out that resolves this.

You may ask is it a testing error? It is in fact not and that is what this section is all about to explain why. First step was rerunning all the benchmarks and system checks with a different DDR5-8000 memory kit. Afterward, it was concluded that a lot of time was wasted, and secondly, something is wrong with the G.SKILL XMP profile. It wasn't entirely apparent why such a large gap exists between other 8000 MT/s kits tested. Next step was to dump the SPD and check if it matched what the BIOS was reporting for the XMP profile. It came back the same. Now onto the troubleshooting. The first guess was related to the CAS value. The simple act of lowering the CAS (CL) value to 38 from 40 gave little to no improvement overall outside synthetic benchmarks, which didn't explain the performance loss. Next attempt was to simply apply the full XMP profile of another 8000 MT/s memory kit. If the benchmarks gave the same results, that would point to a system issue. However, results came back as expected that matched previous tested 8000 MT/s memory kit. This concluded the data collection. We found that the real culprit, and it was these refresh tRFC2 and tRFC-SB values in the XMP profile. For those unaware, the XMP profile applies the primary timings, operational voltage and a few secondary values as well, including tRFC.

These tRFC2 and tRFC-SB values in the simplest description determines in clocks cycles the time that must elapse during a refresh cycle. Essentially, the DRAM cannot be accessed while the refresh of memory cells is ongoing. With tRFC-SB being used for the "Same Bank" refresh, in which the same rules apply. In short, the higher these values are, the "longer" time (clock cycles) is spent being inaccessible to the system. This interrupt must happen due to the physical properties of the memory cells. Also known as volatile storage, it will lose the data in the event of power loss. If not periodically refreshed to ensure that it remains accurate, data bits are corrupted. On the flip side, if you set the value too low, the memory doesn't have enough time to fully refresh, resulting in data loss as well.

After all that investigation, by changing tRFC2 (880) - tRFCSB (760) in the XMP profile to 640–520, benchmarks results came back as expected from previous DDR5-8000 kits tested from KLEVV, Team Group and Patriot, that all hold similar results. This is a 27% decrease in clock cycles wasted on cell refreshing. Below are the results reflecting that change to the CAS and tRFC values. Keep in mind, the voltage for this particular kit did not have to be adjusted from 1.35 V. This is a big win for G.SKILL because all the other DDR5-8000 kits have 1.45 V applied in the XMP profile.

Baldur's Gate 3 Results

Counter-Strike 2 Results

Remnant 2 Results