Let’s first look at the performance in synthetic benchmarks, the AIDA64 Cache and Memory Benchmark in version 6.32.5400 and Geekbench 3 Multi-Core Memory Score 3.4.4 64-bit. All tests were run 3 times and then the average of the results was taken.
In the AIDA64 Read and Write, the overclocked settings on the Intel platform storm away far ahead, as expected, simply due to the significantly higher clock rate. In the comparison of the XMP profiles, the AMD platform is slightly ahead, which will be a trend in the following benchmarks.
AIDA64 Copy effectively means multi-tasking and thus an advantage for configurations with multiple ranks, as it allows the theoretical bandwidth to be used more efficiently at the same clock rate. The dual-rank configuration on the Intel platform can even almost make up the 267 MHz effective clock deficit to the single-rank configuration. However, the triple-rank configuration is on par with the dual-rank setup within measurement tolerance, on the AMD platform that is.
In latency tests, as usual, the architecture-related differences between the platforms are largely apparent. Furthermore, it can be seen that, especially in XMP, multiple ranks first mean more latency unless you optimize the timings.
While the XMP profiles of both platforms were still on par with each other in Geekbench 3, the overclocks on the Intel platform can achieve an equally higher result due to its significantly higher clock rate. While Dual-Rank still has a big advantage over Single-Rank, the difference is negligible on the AMD platform, where the large L3 cache can cover most instructions.
42 Antworten
Kommentar
Lade neue Kommentare
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Alle Kommentare lesen unter igor´sLAB Community →