'Windows 10 Scheduler is not the issue'

no, but you are

File: 1487077561341.png (305KB, 799x634px) Image search: [Google]

305KB, 799x634px

>Arch Lunix: 27138 points
>Win 10: 22722 points
What about Gentoo?

>>59457767
same shit all over again..

amd was talking about the smt moron

windows are still throwing threads around the ccx's which result on a massive latency between the ccx because the threads are using the inifnity fabric which they were suppose to

File: once again v shows its technical understanding of hardware.jpg (542KB, 1024x1275px) Image search: [Google]

542KB, 1024x1275px

>>59459611

/g/ told me that Intel has never had release issues and that only AMD fucks it up with new designs. Clearly you are flatout wrong as /g/ would never lie to me.

File: 1489779080620[1].jpg (19KB, 140x140px) Image search: [Google]

19KB, 140x140px

>Say my scheduler is bad again and you'll find yourself on the bottom of Redmond harbor in a pair of cement shoes

They know that there are issues. They also know that Win10 is the latest meme OS that normalfags want and they don't want to scare them away from Ryzen and will most likely have MS patch it later.

File: 1489465333280.jpg (270KB, 1920x1080px) Image search: [Google]

270KB, 1920x1080px

>>59457767
>Are AMD literally retarded?
Yes.

>>59457767
No, it's clear that the Window schedulers (especially 10) are not ideal for Ryzen, but it's not really an OS bug, and AMD would be suicidal to try to throw a critical business partner like MS under the bus.

It enjoys getting revenue from Xbones and getting driver support, etc., so they can just spew BS until a patch comes along that suppresses thread bouncing and magically boosts performance ~10-15% on the more vulnerable workloads.

>>59457767
HPET and core parking is the issue

File: 1488404508616.png (316KB, 882x758px) Image search: [Google]

316KB, 882x758px

>just wait

>>59462257
>It enjoys getting revenue from Xbones
you misunderstand something ehre, MS depends on them here not other way around
they get peanuts for xbone chips, MS will lose billions in investment if AMD tells them to fuck off

File: 1467296306101.png (91KB, 653x726px) Image search: [Google]

91KB, 653x726px

Fixes are FUCKING coming, o-okay?

>>59457767
Linux results (both ST and MT) are better for Intel too

File: 1486810667607.jpg (66KB, 568x612px) Image search: [Google]

66KB, 568x612px

There's nothing inherently wrong with the architecture itself whatsoever, I s-swear.

>>59462358
not by 25% right?

File: 1464804859737.jpg (52KB, 568x612px) Image search: [Google]

52KB, 568x612px

>>59460147
>>59462257
>>59462296

File: 1488919452885.jpg (514KB, 960x686px) Image search: [Google]

514KB, 960x686px

More cores are worth the less FPS if you gaymen. Go for it goy!

>>59462372
I get 5-6% better ST and 15% better MT results under Linux for my i5-6600

>>59462314
>>59462329
>>59462366
>>59462383
>>59462404
like clockwork. enjoy these free (You)s

File: 1483629357673.png (181KB, 293x397px) Image search: [Google]

181KB, 293x397px

>>59462559
>it is only okay when people hit on Intel

>>59457767
because aggressive core parking is the main problem and not scheduler, you faglord

>>59460179
>1329mhz
>1443mhz
kike

File: amd.png (1MB, 1228x1502px) Image search: [Google]

1MB, 1228x1502px

>>59462314
>>59462329
>>59462366
>>59462383
>>59462559

DELETE THIS RIGHT FUCKING NOW

I HAVE MY LIFE SAVINGS INVESTED IN AMD

>>59459683
wtf I love our boy bill now

>>59457767
When and where did they say that? Need a source, faggot.

>>59459683
This is it, though.

AMD can't just throw MSFT under the bus since they want them to buy Naples too.

Also, apparently more games do their own thread scheduling than most people realized. It turns out that tons of games are just thinking Ryzen is a Bulldozer chip and optimizing as if it's Bulldozer when they'd be better of treating it like Broadwell-E. Lots of them are also similarly seeing them as having as many cores as they have threads, and not identifying them as having SMT. They also aren't identifying the cache structure.

Windows leaving it up to developers to identify CPU topology instead of having a simple and up to date API to poll was a fucking mistake.

File: letdown.png (609KB, 854x640px) Image search: [Google]

609KB, 854x640px

>>59462965
>give it time

>>59462965
Windows was a mistake.

>>59463037
>just wait for Coffeelake

>>59463538
But... I don't have to. : )

>>59463596
I wish I could keep track of you to see how much you regret you 4 cores in 2 years.

If you're a 60fps gaymer you'd have been fine with a 2core, so no big deal. But if you really bought into the overpriced Intel 4 core meme with a 144hz monitor you're going to be regretting it so much.

>>59463681
Because I'm totally an absolute Intel shill the uber Jew told me in my dream it won't happen this decade. ;^]

>>59459623
heh

intel still havent fixed many of the x99 platform issues...

and im not even gonna talk about nehalems and first gen skylakes