http://www.anandtech.com/show/10610 /intel-announces-7th-gen

AMDPAJEETS ON SUICIDE WATCH

This was confirmed months ago and was demoed earlier this month already, this isn't a surprise to anyone keeping abreast of the industry.

>>56338073
why would you be happy for no competition?

>>56338175
>>56338192
>pane;l
intel confirmed finished

>>56338073
1-2% IPC

Kek

I love monopoly

>>56338107
He likes to get fucked by the giant dick that is Intel.

>>56338286
>1-2% IPC
>Kek
I dont understand this, we have known for years intel is working on optimizing power efficiency and integrated graphics performance, NOT IPC gains, yet every year we get people like you acting like intel promised 15%+ IPC gains every year and you're being cheated somehow because they aren't delivering.

They never promised shit, we obviously don't need greater IPC, if the market demanded it, we would have it, that's how it works.

Big data centers however don't need HUGE IPC, they need power per square foot of data center space, the more they can reduce the power and heat the more money they save in the long term.

If Amazon/netflix/ebay/google/etc suddenly NEEDED massive IPC on a single core, you can be pretty damn sure in 2-4 years you'd see massive IPC gains again from intel with little improvement in heat/power draw.

>>56338286
We need smaller power consumption for laptops.
Intel also wants to get into the mobile market.

Nobody needs IPC.

>>56338107
fanboys are like that. they don't work on logic or rationale, but rather on emotional responses to the actions of their favorite or most hated company. they have formed an one-way emotional relationship with a company that does not know nor care about their existence. in essence, they shill and spread fud, and they do it for free like the janitors on 4chan. this applies to amdrones as well as intelfags.

>>56338376
Nope, only intel cucks.

>>56338395
wanna guess how I know you're an amd fanboy (or troll)?

>>56338395
If you dont see that both sides are fucking retarded cucks you're part of the problem friendo.

>>56338107
AMDfags are already supporting that
>Support shitty product
>Keep conpany that should be bankrupt on life support
>they keep producing absolute shit

who gives a fuck, gimme a faster processor than the i5 2500k already

>>56338499
each generation is 5-10% faster, so the 7600k should be ~25-40% faster than a 2500k.

>>56338332
which is the reason the desktop-market is dying.

no home-user really cares about saving a couple Wh.

>>56338640
> 40%
That's nothing, 5 years have passed.
> reeeeeeeeeeeeeeeeeeeeeee

>>56338640
http://cpuboss.com/cpus/Intel-Core-i5-6500-vs-Intel-Core-i5-2500K
toplyl

This means I can watch 4k anime without my CPU going to 100%?

>>56338779
I said
>the 7600k should be ~25-40% faster than a 2500k.

The 6600k should be ~20-35% faster
Here is the 2500k vs the 6600k
>http://www.anandtech.com/bench/product/1544?vs=288
Oh look, it's generally about 20% faster and even 40-50% in some cases.

We get it, you want more IPC because of ur games, well too bad i guess, you wont get it from AMD this time around and intel doesnt give a fuck.

>>56338061
>kaby lake only works with wangblows 10
HAHAHAHAHAHAHAHHAHA
Intelfags on suicide watch!

>>56338061
>Will only officially support Windows 10


What happens if you try to install an older windows version on it? Will they just not have chipset drivers for older Windows OSes available?

What about Linux?

>no native "USB 3.1 Generation 2"-support
TOP
L
O
L

Blunder of the Year

>>56339627
And no HDMI 2.0 either!?
What's the point?

>>56338762
>5 years
it'll probably be 6 by the time it's actually out

>>56339429
Linux will support it probably, but older Windows won't. Unless you want to risk installing 3rd party Russian drivers.
>>56338755
This is false. Anyone who isn't a complete idiot will want to save power.

>>56338889
average framerate increase for the 6700K over the 6700K with both clocked at 4.5GHz is ~13%

It's pretty fucking dire

File: bigfood.jpg (75KB, 634x470px) Image search: [Google]

75KB, 634x470px

>>56338332
>power per square foot
Holy shit dude, are you bigfoot?

>>56338073
They already had this shit since Carrizo

>>56338217
Have you seen their HPC slides? They look likey they were made by children.

>>56338779
>
GeekBench 3 (AES single core)
Data courtesy Primate Labs
Core i5 6500
5,100 MB/s
Core i5 2500K
2,530,000 MB/s

HOLY SHIT IT'S 500 TIMES FASTER.

Now in all seriousness nobody gives a fuck about encryption performance. Even the unaccellerated 5GB/s is already more than enough unless you connect dozens of encrypted SSDs to a single server. 2TB/s may even be enough for dozens of racks filled with HDDs.

>>56339831
I thought power per footballfield is the more commonly used unit.

>>56339840
Carrizo does not support HEVC Main10 hardware decoding, you stupid retard

>>56338755
No, no in home user cares about faster processors. Current gen (and last 4 gens) already do the vast majority of things people need them to do well. There's little need for most people to upgrade their desktops.

Home users are increasingly buying laptops (again, because they do everything they need them to do) for their benefits over desktops. These users do however care about casual gaming (iGPU) and battery life, weight, and heat/noise (processor efficiency).

>>56338061
Quick sync is really the only reason i use Intel cpus. If AMD had a hardware encoder id switch instantly.

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

53KB, 623x1035px

8K HEVC Main10 hardware decoding
8K VP9 10bit hardware decoding

THANK YOU BASED INTEL

Told yall the age of software decoding, encoding was over.

>>56338061
socket??????????????????????

>>56340759
LGA 1151, same as Skylake

>>56340785
Great, thanks

File: forums.anandtech.com_2016-08-30_14-23-35.jpg (298KB, 786x553px) Image search: [Google]

298KB, 786x553px

It's really over for AMD, there's no way Zen can compete. Don't even feel sad anymore.

>>56341044
>igpu

>>56338332
>if the market demanded it, we would have it

The medical market has been demanding immortality at least since the days of the myth of the fountain of youth. You don't get things because you ask hard enough. Why cure cancer when you can make money treating it?

You get R&D investment when the cost of not investing is high enough to justify it. When you're years ahead of your """"""competitor"""""" in IPC you don't invest in IPC.

The reason there's movement in power efficiency is it's easy. After years of "let's just strap a bigger heatsink on it" design there's plenty room for improvement. It's also a benefit of lower nm processes which also leads to using less silicon. So you get to reduce your cost of manufacture, improve your process, and you pick up some enough power reduction to make the switch to ARM not worth it for the customers who value the compute:heat ratio.

Notice how there's competition from ARM in the computer/power arena and intel is making computer/power improvements only.

If AMD suddenly released a processor that COMPETED with Intel, you can be pretty damn sure in a year Intel would blow it out of the water while making just enough computer/heat gains to keep the switch to ARM not worth it.

>>56341196
>50% faster than nvidia 950m

>>56338061
>muh anime decoding
>BASED INTEL

m8 if that's all they have to go on then it's not going to be worth upgrading at all.

File: 7561500_orig[1].jpg (92KB, 1100x581px) Image search: [Google]

92KB, 1100x581px

daily reminder that Intel only uses new litography to jew you out and has been doing that since they rekt AMD.

Literally smaller chips with the same performance (inb4 muh 3% yearly gains per generation) for the same money.

pic related, they do it openly

>>56341510
That's fucking savage.
>here goy, let me show you exactly how we jew you

Got to admit I'm somewhat impressed they released these slides.