EDIT: Changing this comment. For the AES tests, on closer examination of my GB3, GB4 and GB5 scores, the GB3 AES showed use of 33GBps memory, GB4 showed use of 30GBps, and for GB5, I originally was only getting 12.8GBps, now getting around 25GBps after tweaking some interleaving. But, I have to wonder, since they WEIGHTED the performance of that so heavily, literally giving me a 2000 point boost on my score, what the hell are they doing and are they purposely overweighting some aspects of this benchmark in order to favor Intel, or is it to try to balance it with this data point, or what is going on.
John on 04 Sep, 2019 07:04 PM
The Memory Bandwidth and Memory Latency benchmark tests will not be re-introduced. However, we are investigating whether we can introduce memory bandwidth and memory latency as components in system information.
we are investigating whether we can introduce memory bandwidth and memory latency as components in system information
That would be great (may be it is not typical, however, memory bandwidth/latency are the most important metrics I was using geekbench4 for). And I don't care that they will not influence the final score directly.