User Benchmark, a known portal that compares the performance of thousands of processors and with the inclusion of 10 million visits per monthhas been charged manipulate the scores the processor throws to support Intel.
This translates into that. Processors with higher mononuclear performance are applied above multicore performance, something that does not resemble a performance test in a real environment and is clearly harmful for AMD Ryzen 3000 processors with CPU up to 12 core in the market.
Manipulation of this data has exploded in your face, and this can only be seen by Intel Core i3-8100, processor 140 euros with 4 cores that only have higher frequencies has defeated Intel Core i9-9980XE 18-core with prices 2,280 euros In the general classification, even an 'Eloi Gameno' knows that something strange is happening here.
Previously, Userbenchmark accounted for single core performance as 40% of the scorein 4 core 50% and multicore 10%. However, due to the "unrealistic" value of multi-core CPUs such as Ryzen 3000, Userbenchmark changed its rating system to 40% single core, 58% quad core and 2% multicore.
Although Userbenchmark did not express it directly, I could feel it CPU performance in one core is more relevant than performance in many cores. Although the performance of a single thread is still important, it certainly does not become more important than the performance of several cores. The game becomes more friendly with many cores., and when the CPU frequency increases more slowly and finally recedes in the frequency decrease, it will be difficult to increase the performance of a thread, which makes Multicore performance is a more reliable way to speed up the processor.
Userbenchmark finished its response showing that No one complains that the Radeon RX 5700 GPU series is "too high" and warn about "organized army that pumps one brand or another based on favoritism"Button.
Before this accusation, UserBenchmark officially pronounced:
"After the new CPU battery was launched in July 2019, we noticed that our CPU indexes for gaming and desktop computers exaggerated all CPUs with a core number greater than 8, so we updated the index. We estimate that our index has been updated approximate accuracy 8% over the entire spectrum of more than 8500 CPUs.
There is a stir in certain online communities with accusations of favoritism.. These are some real results that emerged from the last update: Ryzen 3000's effective speed rating is affected as follows: Ryzen 3900X is inactive 2 place 3800X +7; 3600X +14; 3600 +13. On the other hand, AMD Threadripper CPUs are degraded and Intel 9900K ranks first in the gaming segment, above the previous rank that occupies the 4th place7 "
Userbenchmark still supports the change, but acknowledges the criticism and says: "We often adjust our effective speed index and hope to add Octa-Core components to the index in time. Changing the 2-core 64% score will be equivalent, up to a maximum score of 16% in Octa-Core components"That would better reflect the performance gains associated with the eight-core CPUs. The company also eliminates very strong warnings about"Organized army shills", something like a troll.
via: Toms Hardware