UserBenchmark calls Media "incompetent smearers" in the face of CPU-weighting criticisms
This is not how you respond to criticism
Published: 21st November 2019 | Source: UserBenchmark |
UserBenchmark calls Media "incompetent smearers" in the face of CPU-weighting criticisms
Criticism is something that reviewers have to live with. With every viewpoint comes its detractors, that's a fact of life. Nobody can be right about everything, and when it comes to opinions, there is never going to be an answer that satisfies everyone.
Over the past few months, UserBenchmark has faced a lot of criticism; both from the tech press and from the PC hardware community. This critique came thanks to the website's adjustments of their benchmark score weighting system, which pushed high core count processors down its official rankings.
The conflict here is simple, UserBenchmark believes that faster single-threaded performance and performance at intermediate core counts matters more than the performance of heavily multi-threaded workloads. For this reason, UserBenchmark has reduced the influence of highly-threaded workloads within its benchmarking database.
Most recently, UserBenchmark has reportedly removed, Multi-Threaded scores for over eight threads from its "Average User Bench", at least according to recent chatter on Reddit.
What's more startling is how UserBenchmark has responded to the criticisms of the tech media. @KeithPlaysPC over on Twitter has spotted additions to UserBenchmark's "About Us" page which calls members of the tech press "incompetent smearers", a statement that's directly linked to the Hardware Unboxed YouTube channel.
As a matter of record, we have saved UserBenchmark's current "About Us" page using the Internet Archive's Wayback Machine. Follow this link to see what UserBenchmark has to say about its critics.
Personally, I think it is a mistake for UserBenchmark to focus so heavily on CPU performance at low core/thread counts. Heavily multithreaded processors are becoming more and more common, and with that shift, today's software is becoming increasingly multi-threaded. Then a 16-thread Ryzen 7 2700 is available for £140 new; it's hard to argue that heavily multi-threaded processors aren't mainstream.
Over the past few months, UserBenchmark has faced a lot of criticism; both from the tech press and from the PC hardware community. This critique came thanks to the website's adjustments of their benchmark score weighting system, which pushed high core count processors down its official rankings.
The conflict here is simple, UserBenchmark believes that faster single-threaded performance and performance at intermediate core counts matters more than the performance of heavily multi-threaded workloads. For this reason, UserBenchmark has reduced the influence of highly-threaded workloads within its benchmarking database.
Most recently, UserBenchmark has reportedly removed, Multi-Threaded scores for over eight threads from its "Average User Bench", at least according to recent chatter on Reddit.
What's more startling is how UserBenchmark has responded to the criticisms of the tech media. @KeithPlaysPC over on Twitter has spotted additions to UserBenchmark's "About Us" page which calls members of the tech press "incompetent smearers", a statement that's directly linked to the Hardware Unboxed YouTube channel.
As a matter of record, we have saved UserBenchmark's current "About Us" page using the Internet Archive's Wayback Machine. Follow this link to see what UserBenchmark has to say about its critics.
Personally, I think it is a mistake for UserBenchmark to focus so heavily on CPU performance at low core/thread counts. Heavily multithreaded processors are becoming more and more common, and with that shift, today's software is becoming increasingly multi-threaded. Then a 16-thread Ryzen 7 2700 is available for £140 new; it's hard to argue that heavily multi-threaded processors aren't mainstream.
(UserBenchmark calling Hardware Unboxed "incompetent smearers")
You can join the discussion on UserBenchmark calling Hardware Unboxed "incompetent smearers" in the face of CPU-weighting criticisms on the OC3D Forums.
Most Recent Comments
Blatant bias is different to a difference in opinions.
I don't trust a single person who takes money from a company and then is expected to give un-biased opinions on that companies products and the majority of that sites advertising has links to intel or intel based products. What did we expect?Quote
I don't trust a single person who takes money from a company and then is expected to give un-biased opinions on that companies products and the majority of that sites advertising has links to intel or intel based products. What did we expect?Quote
I have been vocal about reviewers being biased towards video rendering and blender results instead of recommending adequate CPUs for different workloads. Userbench went one step further and thrown them all into the same basket. No one should visit this site. They can moan as much as they like. This is a definition of a click-bait site that has only profit in mind no matter how misleading their content is. They will soon be buried by responses from proper reviewers.
Which CPU is the best? Well it depends on what you do. It should be "What CPU is best for your use case?" Pugetsystems' team has proper approach to the subject. They even organize them by the application that you use and not just workload in general. GN also started stating for whom is reviewed CPU meant for. For some things AMD wins, for others Intel. Gone are the days when 6950X was the best CPU for everything.Quote
Which CPU is the best? Well it depends on what you do. It should be "What CPU is best for your use case?" Pugetsystems' team has proper approach to the subject. They even organize them by the application that you use and not just workload in general. GN also started stating for whom is reviewed CPU meant for. For some things AMD wins, for others Intel. Gone are the days when 6950X was the best CPU for everything.Quote
Bias is across the board though, was really noticeable when AMD 5700XT was released with some reviewers spending more time talking about RTX than the performance level of AMDQuote
Jokes on them. All reviewers have to do is get together and agree to ignore their software. Less and less publicity, less money, then they cave and apologize, but nobody cares.Quote
CSS Valid
The people who run UserBenchmark are a load of deluded ****ers. They still believe 4C8T is all we need yet in the real world, applications are utilizing up to 32C64T. The world of enthusiast computing isn't just single threaded games, never is, and never was.
It really does grind my gears when software designers pull underhanded stunts like this which in turn makes a community worse since it breeds toxicity. As much as I'd love to say boycott stuff like this, we all know there is a huge circlejerk who won't, and they're the loudest ones and get the biggest platform to scream about it!Quote