http://www.pcworld.com/article/3127 250/hardware/intel-kaby-

>>56942842
Shit looks gay as fuck. AMD all the way right gais? I don't want to get jewed (haha making fun of races is funny) by intel.

>>56942842
How about the desktop i3? Is it finally faster than my i7 860?

Intel made huge IPC gains again, will be more huge for the new desktop prossors

>>56942924
No they didn't. Its extremely marginal.
Kaby Lake is updated media accelerators and tweaked process pushing clocks a bit higher.
Haswell to Broadwell was about a 3% IPC uplift.
Broadwell to Skylake was about a 3% IPC uplift.
Skylake to Kaby Lake is still about a 3% uplift.

>>56942842
>>56942924
>IPC gains
6200U stock clocks at 2.3ghz and turbos to 2.8ghz when single core, 2.7ghz when multicore

7200U stock clocks at 2.5ghz and turbos to 3.1ghz on both.

320/286 = 11.8% ; 3.1/2.7 = 14.8%

It's actually a decrease in that bench lol, probably didn't turbo to 3.1

>>56943029
This.
They just increased the clock on skylake. Thats probably why the graph shows i5. Probably the overclock margin on i7s were smaller so there was practically no gains between two generations

>>56943114
The clocks are actually identical to the i7 6500U

>>56943137
My xps 13 with i7 6500u got 322 points so yeah it seems they just overclocked the skylake i5 to i7 speeds and the gains on the graph are that.

>>56942879
this right here. epic post.
I dont care if Im getting literally half the performance at double the power usage with AMD, they the good boys, they dindu nuffin. Nvidia and Intel are big bad JEWS!

>>56942875
Holy shit that's amazing for laptops.

>>56942924
>will be more huge for the new desktop prossors
should we tell him? or just let him be happy?

>>56943767
The feature was intended to be available on Skylake but had to be disabled due to malfunctioning. Honestly don't you keep up with the spec sheet? It's just marketing hype on features they should have delivered two years back.

>>56942993
Kabylake uses the exact same uarch as Skylake, it's just fab process improvements. If there are any perceived changes to IPC it's most likely testing error or the average turbo not being recorded properly.

File: cinebench.png (12KB, 374x171px) Image search: [Google]

12KB, 374x171px

>>56942842
300 points? are they running this shit on their smart phones?

>>56943767
How many HEVC videos do you watch daily without gpu acceleration?
This is not a general power consumption test, this scenario only applies to h.265 videos that are not accelerated by gpu.

>>56943759
>BUTTMAD AYYMDPOORFAGS DETECTED

>>56943932
wow
your 95W cpu gets 4 times the performance of a 15W cpu
way to go man

>more intel graphs that don't mean shit
CPUs stopped growing in power years ago. Getting any decent fucking processor from anyone will get you almost the exact same performance.

I really wish x86/x86_64 would fucking die already.

>>56943964
>How many HEVC videos do you watch daily

Pretty much an endless loop, actually. I have a 24/7 Plex stream on my HTPC for general background going from my home server which I have slowly been encoding entirely to HEVC. Knocked out almost half of my file sizes across my 12TB of original movies and TV series.

File: 1475580940102.png (20KB, 506x606px) Image search: [Google]

20KB, 506x606px

>>56942842
>>56942875
>higher base frequency and better dsp for video decode

Not seeing what's so impressive here :^)

>>56944048
>HTPC
Thats the exact opposite of laptop though
Its stationary and plugged in all the time
So less power consumption wont make much of a difference
If you have 7/24 stream going on in the background you probably dont care about power consumption anyways
However new decoder means you can watch 8k videos without additional gpu so your next htpc could be a lot smaller

File: dxva_kaby.png (53KB, 623x1035px) Image search: [Google]

53KB, 623x1035px

Why have you not joined the Kaby Lake master race, /g/?

>>56944140
>lock to windows 10

fuck that shit