>most reviewers benchmark CPUs with gaymez
>everyone is suddenly acting surprised that CPUs with more cores and lower clocks perform worse than CPUs with less cores and higher clocks in games which aren't properly multithreaded
I really don't get the point of using gaymez for benchmarking since they can't even fully utilize CPUs. Every fag out there is going "dude ryzen DOA" lmao since it gets BTFO in gaymez, by that logic the whole X99 lineup is trash too since they aren't as good as quad cores in gaymez. Gaymurfaggots were a mistake.
>inb4 ayymd advertised them to gaymurs in the first place
Probably to test the microcode and prepare for the release of naples.
>>59246918
Welp, when your subscribers are a sour, grimy community of hot pocket eaters who can't wait to put their rigs to the test playing Call of Duty, the entire gaming generation takes a step back.
>>59246918
Games are the most popular software with demanding CPU needs. Terrible time to grow up op. Feel sorry for kids like you.
>>59246918
>suddenly acting surprised that CPUs with more cores and lower clocks perform worse than CPUs with less cores and higher clocks in games
honestly i think the average person would assume more cores and threads would equate better game performance
especially since the largest platforms for gaming the xbone and the ps4 all have 8 cores and have been out for years.
>>59246918
>>inb4 ayymd advertised them to gaymurs in the first place
>Probably to test the microcode and prepare for the release of naples.
this is good bait, i think youll do well
should have launched with Opterons
>>59246982
>this is good bait, i think youll do well
Amd simply lacks the funding/budget intel has so it's pretty reasonable for them to use gaymurs as a test subjects for the enterprise market, which arguably bring more cash to them.
>>59246941
>demanding CPU needs
>implying they aren't GPU bottlenecked more
>>59246918
>I really don't get the point of using gaymez for benchmarking since they can't even fully utilize CPUs.
Well, what kind of test you run (or pay attention to) depends on what kind of tasks you want the performance for. If you want performance because you like playing games, games are a more useful benchmark than any other kind of load testing.
>>59246918
Not to mention, the performance is pretty stellar for windows not being patched for SMT yet (apparently windows will see 16 actual cores and not recognize SMT properly).
I was just happy we actually got broadwell-e ipc out of ryzen on their first generation of 14nm/finfet, can't wait to see how optimization goes vs intels optimizations process.
>>59247124
Optimization?
Downgrade to win7 and you'll see 18% fps increase. There's no need for optimization
>>59247183
That says more about Windows 10 being shit than it does about Ryzen's performance
>>59246918
Ryzen isn't just bad at games, it's bad at almost everything
>>59247320
>single threaded benchmark
>>59246918
holy fuck someone saved my picture
>>59246918
>we can't get good single thread performance but we fixed that problem by making more cores
>it's your fault if you can't utilize all of the cores
>>59247320
>overclocked 6900k is worse than 1800x
topkek, rip $500
>>59247022
>use gaymurs as a test subjects for the enterprise market
Isn't it a bit kind of irresponsible act in a customer's perspective especially it's AMD's (and ALL of the CPU designers and manufacturers') job to do validation? I mean, they have literally 4 years to prep the CPU. Leaving the abandoned objectives like dual-x86+ARM-architecture-in-one-chip thing aside, let the gaymurs and potential just-want-to-upgrade-an-old-PC customers fall into the "trap" and help testing the shit all day, finding bugs and reporting to AMD and folks...
Maybe those bugs will not be as serious and sudden as FDIV and prime95 bugs on Intel CPUs, but still.
>>59247320
It's slightly ahead of the 6900k, which is it's main competitor.
>>59247395
intel's more cores option is also worse than the ones with lower core count. You can't beat physics, cunt.
Also, only windows software is bad with high core counts. There is a reason why only linux is used on servers. Have fun waiting for patches to play catch up.
>>59247570
They probably have performed testing on their own. Certain uefi bugs, minor glitches in the microcode might not be anything of importance for an average consumer but they are certainly important in an enterprise environment where uptime is the main objective. For instance I'm sure we'll see major uefi updates for current AM4 boards very soon.
Buying hardware just after its release is usually not a reasonable thing to do regardless of the company.