Coffee Lake 11% Faster IPC

>Intel claims
>claims
nice try there, shill

>>61950404
If it's clock speeds are 3.8 base, 4.3 all core boost and 4.5 single core boost, does this pretty much guarantee it's soldered?

File: AMD fucking finished.jpg (180KB, 935x923px) Image search: [Google]

180KB, 935x923px

>>61950443
Nice try AMD shill, you've already had two disasters this year (Vega and Threadripper), Intel is about to wipe out Ryzen and make it a straight three for three disasters for AMD this year.

File: Capture.png (7KB, 469x250px) Image search: [Google]

7KB, 469x250px

>>61950466
>>61950404

ANON! I just got this insane leak from AyyMD! Cand you believe it?!

>>61950404
based

File: 10082017114652.jpg (638KB, 720x3689px) Image search: [Google]

638KB, 720x3689px

>>61950466
>gaymen benchmarks on a HEDT platform nobody uses for gaymem in the first place
I really really love this meme

File: 1502975561525.jpg (267KB, 1400x789px) Image search: [Google]

267KB, 1400x789px

>>61950404
1. Already a thread for this trash >>61947076
2. Believing Intel marketing slides puts you around the level of a complete retard in terms of intellect

File: 1502413488795.png (83KB, 980x735px) Image search: [Google]

83KB, 980x735px

>>61950466
https://arstechnica.com/gadgets/2017/08/amd-threadripper-review-1950x-1920x/

Nigga, 3 out of 4 processors i use on a daily basis are intel. Thread ripper and Ryzen are both fantastic chips. If i was building a new system today to replace my main one It would be based on a ryzen 7 platform.

>>61950404
>11% Faster IPC
You really believe this shit, I've got a bridge to sell you.

>>61951338
You really think Intel is just going to make things up on their slides?

>>61951338
11% from nothing is still nothing
Itnel just make this up for 3 generations
Broadwell ---0%--->Sky Lake---0%--->Kaby Lake---0%--->Covfefe Lake
Eventual 0~2% performance bumps bumps for higher clock or magic instructions

...Wait
There was slight degradation of performance from Broadwell to Sky Lake, thats why itnel disowned 5th gen and banned it in comparisons to skylakeu and further gens

>>61951338
> >11% Faster IPC
No, 11% higher performance. Clocks are higher.

>>61950443
>amd claims
IT'S TRUE GOYIM
>intel calims
REEEEE IT'S FALSE NEWS

File: konosuba2.jpg (205KB, 1920x1080px) Image search: [Google]

205KB, 1920x1080px

>tfw OP recalls that intel said skylake was going to be 20% faster than haslel

>>61950484
>(lower is better)

good thing about the drop in clock speed it really has little effect in games, a 5.1 7700k gets 1-2 more fps then stock the only benefit is in the lows witch adding 2 more cores would help much more then a few 100mghz

if there is no mesh or kaby lake TIM issues with a clock speed of even 4.5-4.8 it should blow the doors off ryzen in games

File: f58908fbc9096bb4bb995de242433bf5.gif (1MB, 400x225px) Image search: [Google]

1MB, 400x225px

>>61951557
>and that kaby was 10% faster than skylake

>>61951542
>>61951542
Intel lies on every one of it's slides about performance improvements using fudged Sysmark comparisons.

>>61951557
>>61951585
icecake confirmed for 0.125% ipc increase over babylake (+30w [125w socket], +brand new socket [for lulz really], +100usd msrp [because it's so awesome and new], +2 cores on models above 400 [for +200 usd a pop])

>>61951536
>>61948340

>>61951635
AMD lies too yet /g/ eats it up.

>>61950466
>wojak
KIll yourself my man

>>61951475
your joking right?
G L U E D
L
U
E
D
or should i say SOPA? im conflicted now

File: 1502075488963.png (59KB, 548x562px) Image search: [Google]

59KB, 548x562px

>>61950466
>threadripper
>disastrous

File: lol.gif (2MB, 390x277px) Image search: [Google]

2MB, 390x277px

>>61951475

>>61951673
[Citation Needed]
RTG has shown some fairly stupid stuff (lol 480 SLI) but I don't recall anything of the sort out of AMD.

>>61950466
>X299 disaster
>Metal Xeons DOA due to Epyc
>Coffee Lake won't be a disaster this time we swear!

>>61951736
intel is on a roll, all the the way down.
I really want to check server market share q4/17 q1/18

>>61951716
(gtx) 480 sli and (rx) 480 crossfire are two completely different things anon...

That is nice, Intel would never make some bullshit based on a highly optimized workload right?
Let's see what the 9th gen marketing wank after that will be.

>>61952719
>Every processor you buy saves 1 african child
>Please buy our processors (sad hungry children appears)
>AMD is literally devil

>>61952719
maybe playing a game in vlc? no wait, that was already done

but 10nm CPU in 2018 will be revolutionary right?

is it worth waiting or buying Coffee lake now??

File: kaizad-mistry-2017-manufacturing(1)_29_575px.png (103KB, 678x381px) Image search: [Google]

103KB, 678x381px

>>61953013
According to Intel's own graphs, 10nm+ isn't going to be any faster than 14nm++. Their 10nm is just a mess.

Didn't Intel clain kaby Lake was 10% faster than Skylake? We all know how that turned out.

Remember how Intel claimed that 7900X is 140W TDP?

Just like Kaby Lake was 15% faster than Skylake

woah so 6 core 12 thread i7 at the top end of intel's consumer lineup
crazy shit man

>>61950404
Ryzen is still competitive, and H1 next year we'll see Ryzen tweaked with faster clocks.

>>61953086
either 28% better efficiency or performance over coffee
you should wait

>>61953838
My 2500k is still competitive, so that's not saying much.

>>61953845
Where does it say that?
>>61953906
2600K is the only Sandy worth using in 2017

>>61953931
it goes down from 14nm to 10nm

>>61953953
Yellow dots on the graph on the right are only going towards the side, meaning it'll just take them longer to get there.

>>61953931
No, 2500k is perfectly fine.

File: 2600k-revisit-mll.png (87KB, 806x681px) Image search: [Google]

87KB, 806x681px

>>61953979
No, it isn't.

>>61953974
then I suppose it will be 28% more efficient rather than increased performance

>>61954003
>Well above 60fps
>Only dips lower because it's working full speed to produce more fps
>Somehow that's a problem
>Still perfectly good for 144Hz fps games
>Singleplayer games still run smooth at 60Hz
The 2500k will be older than the people i'm shooting at before it stops being competitive.

>>61954003
Oh come on now, Metro LL is basically cherry picking since it's one of the very few games where i5's get minimums that low. Honestly I think it's a problem with the game as it's not very CPU-intensive, there's nothing to explain the drops in performance. All the other games he tested had more reasonable scaling, the only other one with framerate drops was Total Warhammer but every CPU struggles with that title including Ryzen.

>>61954094
>smooth
>pure garbage lows

Hmm.

>>61954203
It doesn't change the fact that a 2500K overclocked to the limit only matches a stock 2600K. So why would anybody want all that extra heat and power draw, when they could buy a 2600 non-K and enjoy the same performance (with much better lows), with far less heat and power draw?

The 2500K is e-waste at this point. Only the poorfags still desperately clinging to it disagree.

>>61954256
Who are you quoting? And if you try to render a ridiculous amount of frames you will eventually get framerate drops because the CPU can't keep up. Lock framerate to desired rate +1 one pre-rendered, and it'll be smooth sailing. 2500k is capable of 120fps BF1 easily, which is one of the more demanding games where you actually need more than 60fps.

>>61954203
>metro last light being ousted as cherry picking in a scenario that kinda favors AMD.

boy oh boyo, today was a weird day.

>>61953086
what is this chart? No, what the fuck is this chart!
X axis is fucking time? A FUCKING TIME!!!

>>61954003
>OCed 2600k is on par with all the top CPUs
>2600k isn't fine
You need a brain chack.

>>61950404
Just remember to buy a new motherboard goy

File: Metro-Last-Light-Cpu-Benchmark.jpg (106KB, 773x693px) Image search: [Google]

106KB, 773x693px

>>61954330
It's just because it's one of the very few games to show such a drastic difference between an i5 and i7. Crysis 3 likes more threads as well but the difference is no where near as high. Also it could always be his setup as well, or maybe a patch broke something, I found this from a different website.

>>61950404
>IPC
You know that IPC is "per clock"?

Coffee probably has zero core IPC uplift. THey just found some corner case of a benchmark that benefits strongly from the bigger L3 cache or DDR4 speed (2666MHz support instead of 2400).

5% of that boost is higher turbo clock.

>>61954376
>somebody said the 2600K isn't fine

You need an eye test or a bullet to the head.

File: UMA_MACACO_STRIKE_AGAIN.png (98KB, 361x453px) Image search: [Google]

98KB, 361x453px

>>61950404
>BRAZILIAN EDUCATION

File: Capture.png (198KB, 739x900px) Image search: [Google]

198KB, 739x900px

>>61950466
>I love to get paid to post this!
>I love to cherry pick results!
>I love to ignore the stupidity of someone spending 1000$ and assuming they get better performance in games!
>I love gaymen!
>I love being so stupid as to not realize the utility of additional cores!
>I love to ignore actual facts like Ryzen selling more CPUs than Intel!
Here's the full results with explanations in the picture or link: http://www.gamersnexus.net/hwreviews/3015-amd-threadripper-1950x-1920x-review/page-2

>>61951475

http://www.youtube.com/watch?v=TAGrPyMKA_k

>Mooly "shmuel" Eden
>Intel senior vice president