Why do you block global PCI mutex during memory throughput testing?

[1usmus]

Why do you block global PCI mutex during memory throughput testing? 
Why do you always have the cache tested not by the best core, but always by the 0? Perfect pattern for Intel and disgusting for AMD.

[Fiery]

4 hours ago, 1usmus said:

Why do you block global PCI mutex during memory throughput testing? 
Why do you always have the cache tested not by the best core, but always by the 0? Perfect pattern for Intel and disgusting for AMD.

1) Do you mean the Global\\Access_PCI mutex?

2) I'm confused. Do you mean the throughput (bandwidth) benchmark for the caches or the cache latency benchmark?  And what CPU do you think we measure an incorrect performance of and by what margin?

3) Are you the author of Ryzen Tweaker?