What really surprised me about the modern AMD processors was how they do switching the threads, distributing the workload between the cores. Recently i tested it a lot in a different test cases to make sure there's no any mistake in what i see about it... Whatever test case i tried i didn't see any performance losses at all, which was probably the most unexpected and surprising thing of this processor for me, i couldn't get it how is this ever possible.
I didn't use AMD processors since the times they introduced S754 / S939 platforms many years ago, not at home, not at work, nowhere, so wasn't aware what's the actual progress these processors had for all these years.
With Intel processors since the Core 2 series and further it was always normal to have some performance losses on switching, including a Haswell platform mentioned above. For this reason we always had to mess up with affinity settings in cases when it was necessary to get the highest possible performance out of the single threaded applications. It was quite inconvenient, but there was no other choice, as it was practically impossible to redesign those applications.
Now i see AMD processors don't have this problem at all, which i consider is a great achievement and a huge technological breakthrough AMD has. I also don't understand why i didn't see this important property of these processors actively advertised, and so didn't even know about it until i've seen it myself when got these processors for tests.
I don't have exact information if this kind of performance improvement has been introduced only in the latest CPU series by AMD which we have tested, or probably in some earlier series as well.
It is highly likely Intel has some similar improvements introduced in their most modern CPU's too (if somebody has proven information about it, please, write about it below).
___
What about general PC performance - it's being measured not by "feels", but by benchmarks, specialized tests and applications - that gives information about particular units and subsystems performance, and also overall performance that will be achieved in real applications. Also, it's represented in exact number or rating - which usually lets you know how many operations of a particular type a CPU (or another PC part) performs in a particular interval of time. An optimal choice of the CPU model might be different depending on the applications a CPU will be used for.
If there's "feels" one PC works slower than another, there might be huge number of reasons why do this happen, where a CPU performance is only one of them.
For example, one probably may notice all applications and even bat-scripts in Windows do startup incredibly slow. Do this happen because a CPU performance is too slow? Maybe... But, it also could happen because antivirus sends an executable or a script to its servers for testing which may take a while (depending on the servers load) and until the testing is done an executable/script won't be allowed to run.
It may happen because of too high disk fragmentation. It may happen because of Windows pagefile traditional issues - when in some cases, by no actual reason, it moves an applications' data from the RAM to the disk, which reduces their performance a lot.
If we begin to explain each of possible reason that potentially may reduce the PC performance in one or another part, there may appear a thick book...
In case OP described, a tests/benchmarks results have to be compared for both of mentioned PC's, including memory and disk benchmarks. Probably it also worth to check them in a DPC Latency tests. Only these results will show up what part makes the PC to "feel" slower.
I'd suspect a slower one has significantly slower SSD disk or a more heavy software environment (antivirus, some software that utilizes the disk actively - such as file sharing, etc.).
