Winbabby tries to disprove Phoronix

>winbabby tries to disprove Phoronix
>gets the same results
AHAHAHAHAHAHA WINFAGS ON SUICIDE WATCH

Attached: test.png (3844x2164, 1.27M)

Other urls found in this thread:

phoronix.com/scan.php?page=article&item=2990wx-linwin-scale&num=1
twitter.com/NSFWRedditImage

>owning threadripper or a core i9
absolute /v/ cringe

Now I'm really tempted to upgrade to a TR 2900

I thought this has been a known thing since supercomputer all use linux

Brainlet here.

Aren't there special versions of Windows designed speficially for this kind of stuff?

>/v/
Linux is anything but /v/, baby.

*Cries in bash*

stop crying

>threadripper on linux
>/v/
ahahhaha get the fuck out of here user

Jesus, why is windows such a pile of poo?

>by trying to set up Linux I started to appreciate Windows 10 more
Steve confirmed cuck

Jesus fucking Christ that performance difference is colossal.

No, windows scheduler is universally shit.

We have known this since Zen 1. Winblows literally shits the bed on high core counts and throws threads all over the place.

Wasn't per core licensing in the drawer?
aka, how to make professional pay more for no reason?
A wild guess is, M$ gimps high core count CPUs because they need to sell Windows Server licenses.

phoronix.com/scan.php?page=article&item=2990wx-linwin-scale&num=1
Nope. Even the windows server edition is shit compared to desktop oriented Linux.

I switched to Linux a year ago and I've never missed Windows.
Back in the day Windows was more polished and Linux gave you workflow headaches. These days Windows is a pain in the ass and Linux much more polished. Speed is better in Linux, security. The only reason to use Windows is games

it's actually worse on server 2016.

>threadripper
>linux
>>>>>>>>>>>>>>>>>>>>>>>>>>

Attached: 1515405625810.png (675x827, 35K)

Linus Trollvalds says the benchmarks favor Linux and that the I/O starved nature of Windows makes it look worse. I don't buy it.

Attached: Screenshot_20180820_113732.png (888x466, 114K)

>windows 10 based systems are even worse*
i wonder if the old trick of having windows media player running tin the background during the tests still improves the results

wtf is this real

It does not sound unreasonable that the OS takes that as a hint of needing soft-real-time performance

is it windows 10 thing or windows in general

please get the fuck out, stop clogging the board with your consumerist trash

it used to be, at least, dunno if anymore thats why im asking
iirc it made the kernel timer run at a higher rate, which improved performance is some edge cases

Because Microsoft no longer cares about Windows in its current incarnation. Its coasting entirely on its current install base and legacy software library.

They know they don't have the server space, they know they can't contest the home/pro desktop space going forward. Like it or not Windows is morphing into a iOS/Android style OS for media consumption devices and dumb frontends for office cloud services.
They're hoping the Xbox/UWP game library will carry them with young people until they age into the work force and that their lead in the office desktop space will last long enough for that to happen.

Attached: 1532202345361.gif (500x493, 631K)

Bravo pajeets

How come 7980 is faster than 2990? I think there is something wrong with the tests.

The 7980XE clocks higher, has slightly better IPC and has less latency between cores. Also, not all benchmarks scale as well with the number of cores/threads, something Steve refuses to acknowledge despite the glaring evidence he himself gathered.