Op vrijdag 25 januari 2019 11:30:11 CET schreef Vinzenz Vietzke:
Am Freitag, 25. Januar 2019, 10:10:43 CET schrieb Michal Kubecek:
For instance, just few days ago he published an "out-of-the-box 10GbE network benchmark" which is extremely crappy even for Phoronix standards. He completely ignored that some distributions enable connection tracking and set a complex set of netfilter rules while others do not setup any kind of firewall and compares network latency numbers between them. In one "bandwidth test" he reported that the fastest result among all tested distribution was "847 Mbit/s" (through 10Gb/s ethernet) but didn't stop for a minute to say "Hey, something is obviously wrong here." and didn't do so even after two other programs reported numbers more than 10 times higher.
To second that somehow:
I don't have the source at my hands anymore but recently came across a blog post where someone ran Phoronix' "test suite" benchmarks multiple times with the same settings and on the same system. He compared the results and got *vastly* different figures. I mean, there were deviations up to 25%!
That is something to keep in mind even for benchmarks from other sources if they use that test suite...
Regards, vinz. So, let's conclude that Phoronix is a click-bait site. A hardware geek in my town had experiences likewise. Question arises: who's paying the guy?
-- Gertjan Lettink a.k.a. Knurpht openSUSE Board Member openSUSE Forums Team -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org