Intel Core i5 7600k Kaby Lake-S Performance Review & H.265 testing

[HL0315HI]

Call me "Dipper315" instead.

-----------------------Source---------------------

E-Zone Hong Kong Issue 953 11/17/2016

-----------------------Article------------------------

First Page:

Second Page:

Third Page:

Fourth Page:

Fifth Page:

Sixth Page:

Seventh Page:

-------------------------------------------------------

Picture not loading for you? View it here:

http://imgur.com/a/hPq3S

-------------------------------------------------------

Busy handling finals, so I don't have spare time to translate it.

Find Dennis.

 

[iRileyx]

Yay, more Kabyleaks.

[Simon771]

Should we take that with a grain of salt?

[jagdtigger]

i5 6600 600 points in cinebench

i5 7600 669 points in cinebench

i5 4670k@4,5GHz 692 points in cinebench

 

I think i skip this gen too .

[Morgan MLGman]

This is just sad because there are almost no performance improvements in terms of IPC... 7600K is faster in those benchmarks only because of higher stock clock speeds.

[DELTAprime]

It's in line with what we know of Kaby Lake so far. Faster base clock = performance improvement over last gen, other than that little change.

 

I'm still waiting for Kaby Lake. No reason to buy old gen silicon when the new stuff is so close. Besides the Z270 leak indicates more PCIe lanes, I hope that rumor is true.

[TetraSky]

13 minutes ago, jagdtigger said:

i5 6600 600 points in cinebench

i5 7600 669 points in cinebench

i5 4670k@4,5GHz 692 points in cinebench

 

I think i skip this gen too .

So it's yet another year of Intel sitting on their asses doing nothing productive. Alright then.

[unsubscribed]

None of the images display for me. So there's that...

[8-Bit Ninja]

2 minutes ago, TetraSky said:

So it's yet another year of Intel sitting on their asses doing nothing productive. Alright then.

sad part is they don't need to do a thing, in the last four years amd has given nothing that comes close to the i5's and i7's, i hope zen is a good as people say because then Intel will have to get up and do something. 

[TetraSky]

14 minutes ago, 8-Bit Ninja said:

sad part is they don't need to do a thing, in the last four years amd has given nothing that comes close to the i5's and i7's, i hope zen is a good as people say because then Intel will have to get up and do something. 

Yeah I know. That's probably the only thing I'm hopeful about for AMD... But I just know they will fuck it up somehow and even ZEN won't perform as well as the current offering from intels.

[8-Bit Ninja]

Just now, TetraSky said:

Yeah I know. That's probably the only thing I'm hopeful about for AMD... But I just know they will fuck it up somehow and even ZEN won't perform as well as the current offering from intels.

Of Couse it won't, people are hyping up zen to be something its unlikely to be, I just really wish it was, because competition is never a bad thing, and giving intel a reason to be competitive about their products is better for everyone. Instead of the current Intel which is "You don't like our cpu's, get a worse one from amd then"

[Tech_Dreamer]

kinda predictable performance bump, like previous times

[Misanthrope]

Clock seems to get high but I suspect binning the normal bins further. Though if most users can hit 5 or beyond then it'd be interesting.

[TidaLWaveZ]

I wonder if 4k netflix will attract any buyers, seems like the only thing this processor really offers over it's predecessors.

[patrickjp93]

Magnificent 10% boost over Skylake, officially breaking the 100% lift over Vishera clock for clock.

 

As much as everyone loves to complain, they really should do their research. You can't make instructions execute in less than 1 cycle. The problem is software, not Intel.

[randomhkkid]

1 hour ago, patrickjp93 said:

Magnificent 10% boost over Skylake, officially breaking the 100% lift over Vishera clock for clock.

 

As much as everyone loves to complain, they really should do their research. You can't make instructions execute in less than 1 cycle. The problem is software, not Intel.

What you're saying makes no sense, as a Sofware Engineer you should be aware that multiple instructions can be executed per cycle. This differs per CPU architecture, hence the Kaby lake being 2x IPC of Vishera as you mention.

 

Reading: https://en.wikipedia.org/wiki/Instructions_per_cycle