An innocent-looking slide was released by Ryan Shrout of Intel (PCPerspective’s editor-in-chief, introductory film). It presents SSD storage on the Intel platform Rocket Lake (Core i9-11900K) achieved a 11% higher transfer rate in PCMark than on the AMD platform.
More inquisitive users visited the Intel website, where more detailed data was supposed to be given, but the specific measurement result is not given there either – just the claim that the Intel platform came out 1.11 times faster than the AMD platform. This is strange (meaning the absence of measured results), but not essential. The following configuration is more interesting:
Smart users pointed out that the board with which the AMD platform was to be tested, the X570 ROG Rampage VIII from Asus, does not exist. Asus has not really released such a board (at least so far), or there is no board with the X570 chipset in the Rampage series. So we can only speculate on what Intel was actually testing. It is not possible to verify the measured values - Intel did not publish the values, the platform Rocket Lake released and therefore will not be available for over a month and the platform with the X570 ROG Rampage VIII board does not exist.
Maybe Intel really hit the ground: If no one can remeasure it, then we have no way to deny that Rocket Lake is faster. Maybe there will be something to it. At CES, Intel claimed that Rocket Lake is 6% faster in Metro Exodus than “12-core Ryzen 7”. There is also no such product, so so far no one has been able to refute or confirm Intel’s claims.
But that’s not the end of it. Finally, when asked about the configuration, Ryan Shrout stated that the SSD was not tested in the native NVME M.2 slot, but via a PCIe riser, ie a PCIe card with an SSD slot or slots. Apart from the fact that in such a situation about 99% of desktop users do not run an SSD (uses a native slot on the board), there is also a lack of information on what the PCIe riser was about. There are both passive and active solutions with their own controller. Again, we can only debate what Intel probably used and why it didn’t put the tested SSD directly on the motherboard at all.