https://forums.anandtech.com/thread s/intel-skylake-kaby-lak

Good god such low base clocks for the 10 core, that leaked ES 16 core is 3.1GHz and that's still not even a QS

>>60344618
>Core i9-7820X
>28 PCIe Lanes
This is painfully disappointing. Fucking 8C CPU and it still has so few PCIe lanes? I didn't buy Ryzen and it looks like I'm not buying this shit either, since the 10C model will likely go for $1000+ and I'm not paying that much.

>>60344656
It's just abortion Xeon chips, there are Skylake Xeons with 12 cores at 3.6GHz base
Though probably 180W TDP

>>60344618
>Core i9

I'll believe it when I see it. Intel has never used "i9" for anything so this is probably just made-up bullshit on a random forum post.

>>60344618
Not that SLI/Crossfire is something that should be recommended, it'd still be nice to have the option of 2 cards at a full 16x without dropping over $1000 for a 10C/20T monster CPU. Seems like a step backwards from even the 6850k.

>>60344741
Skylake-EP goes up to 205W so they can go crazy on clocks.

>>60344752
>Seems like a step backwards from even the 6850k.
It is, unless they're somehow pricing that 10C CPU at 6850k levels, which will never happen.

>>60344618
Fuck, if these are the clocks on the "Enthusiast" 6+ cores, what are they going to look like on the consumer Coffee Lake-S? Probably like dogshit.

Looks interesting but I'm not seeing the benefit of cutting the L3 cache to one third and replacing it with L2 since the L2 isn't shared

>>60344618
>Core i7-7640K
>4C/4T
>on a HEDT platform
This chip deserves to be the butt of all jokes for years to come.

>>60344832
Gimped i7 on HEDT, thank you Intel you are truly our greatest ally.

>>60344618
M O A R C O A R Z
O
A
R

C
O
A
R
Z

Core i9-7900X = $1000
Core i9-7820X = $800
Core i9-7800X = $600

>>60344821
In a few hundred years someone will optimize software for 1mb L2.

>>60344821
No major benefit, large pools of shared L3 weren't producing good yields.
There would be a benefit if applications take advantage of the 1MB L2, but everything in the last 10 years+ has been centered around 256KB L2 that's a far way from happening

File: 1485383818384.png (912KB, 1110x1080px) Image search: [Google]

912KB, 1110x1080px

>>60344862
>Core i7-8700K - $400

>>60344794
>205W
Jesus christ, even Naples isn't going over 180W and that's 32 cores at around 2.7-2.9GHz clocks

>>60344820
The consumer platform always had higher clocks than HEDT though.

File: 1489122339460.png (495KB, 1070x601px) Image search: [Google]

495KB, 1070x601px

>>60344924
IT'S NOT FAIR

>>60344971
Isn't that because HEDT platforms had moar coars? Coffee Lake-S is 6 cores mainstream for the first time.

Unless the all core turbos are a good 600MHz over their base clocks, which is a good fucking joke as Intel's TDP is based on their base clock, I don't see how these will be a good buy unless they're $300-700

7820X Would need to cost under $600 to compete with the r7

>>60344971
Because 4 core limit, Coffelake-S is 6 cores mainstream and that platform needs to be under 95W

Can you imagine how low the clocks will be?
We already got sisoft leaks of the 6 core at 3.5GHz, that's a fucking QS

File: gabudoesnotcare.png (529KB, 915x692px) Image search: [Google]

529KB, 915x692px

>>60345005
>Intel
>Competing on price

>4.5 ghz 6c/12t with kabylake IPC
If they sell this anywhere near 1600x prices Ryzen is dead

>>60345020
Yes, in fact let's link it http://wccftech.com/intel-coffee-lake-desktop-6-core-4-core-cpu-leaked/

>>60345029
Nvm read that wrong, only 4 ghz on the 6c wtf

>>60345029
Ryzen has higher IPC though.

> - 112W for Kaby Lake-X / up to 160W for Skylake-X
>up to 160W

Haha oh wow, all their previous HEDT was 140W
They're literally trying to push with more power usage.

n e t b u r s t

Literally D O A housefires.

>>60344618
>up to 160W
Should have gone with 200W, Intel, you might have gotten 100MHz more out of that turbo clock.
Lmao

I wonder how high will AMD's HEDT shit clock.

>>60345029
>he doesn't understand how turbo works

>>60345029
>>60345061
Not only did you read it wrong, but the idea Intel would sell their shiny new 6 cores for that much less than their old 4 cores is ridiculous. They never lower their goddamn prices.

>>60345108
There are talks they want to get the 16 core to 3.6 base clocks.
They'll probably hit 3.5GHz at best, which is still amazing.

>>60345108
Leaky leaks http://wccftech.com/amd-ryzen-16-12-core-x399-whitehaven-cpus-leaked/

>>60345144
That and 64 PCI-E lanes means that Skylake-X is DOA.

>>60345116
I imagine it would have to be more expensive than a 7700k, but not by alot, else people will just go for Ryzen

>>60345144
Shouldn't be a problem. They just need two Ryzen does that will hit the desired clocks to stick on an interposer, really. The bigger issue is the motherboard's power delivery and cooling.

>>60345169
>64 PCI-E
That's free PCIe lanes, once the mobo takes its share of M2/U2/SATA/whatever you'll be left with 52+

well there goes any hope for AMD ;_;

>>60345188
I expect the i7-8700K 6 core to sell for $375-$400.

>>60345196
Still more lanes and more coarz than SKL-X.

>>60344618
so i9 is a xeon replacement? are they just a plethora of bins now?

>>60345205
Then the question will be, does it OC past 4.5 ghz?

Judging by the boosts on the larger chips, it will.

If it does it might be worth that kind of cash.

>>60345230
Are you ready to delid & watercool it?

>>60345241
Delid yes, watercool no, proper aircooling will suffice for a delidded chip

>>60345251
not when it's in 300w range at those clocks.

>>60345251
No, you'll need to watercool it, see Broadwell-E.

>>60345214
I doubt they're real
If they are, 100% chance its a way to offload a bazillion gimped xeon chips

>>60345268
>>60345277
I will wait for the actual benchmarks to see if that is true

>>60345230
>does it OC past 4.5 ghz?
You're just praying you win the silicon lottery at that point.

>>60345297
It's 6 core Intel chip. It's HOT. It has less L3 cache than its HEDT brothers so it'll be slower. tl;dr get 1600 and stop being a faggot.

>>60345308
He should be praying his mobo can handle 300W from the socket and that magic smoke doesn't appear.

File: come_on_jews.gif (2MB, 375x274px) Image search: [Google]

2MB, 375x274px

What the FUCK, are you kidding me? They completely gimped the 6850k and 6900k equivalents with this 28 PCIe lane bullshit. Now I would have to splurge for a 10-core chip to get more lanes than that, and what are the chances that shit is below $800?

AMD better deliver with their HEDT shit, goddamn I'm disappointed.

>>60345314
a 1600 would be a significant drop in core speed for me, and literally be a sidegrade at most.

I need a cpu with at least 4.5 ghz on the clock or it's not worth upgrading

>>60345308
Since they are releasing the larger chips with a 4.5 ghz boost out of the box, the 6c one will hit it even more easily

>>60345322
Or you wait for Pinnacle Ridge with actual IPC improvements instead of moarghz housefires?

>>60345344
Or I wait for actual benchmarks instead of dank memes

>>60345353
>implying it's not another set of damage control housefires
Healthy kek.

>>60345373
ok kid

>>60345353
What memes? Pinnacle ridge or Skylake-X which is at higher TDP than Broadwell-E was?

>>60345338
You're hoping for too much out of a binned to shit consumer 6 core CPU.

>>60345314
>6 core Intel chip
>$400
>R5 1600
>$219

lol, the 1800X beats even Intel's 10 core CPU

>>60345435
We shall see

>>60344618
J U S T krzanich my shit up please!

>>60345319
I'm disappointed too, but I highly doubt a 12C AMD chip is going to end up less than the $700-800 area, considering how their top 8C ends up at $500 already.

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

316KB, 882x758px

>>60344618
IT'S NOT FAIR AMDBROS

File: JUST krzanich.png (1MB, 1002x1480px) Image search: [Google]

1MB, 1002x1480px

>>60345469

File: photo[1].jpg (50KB, 900x900px) Image search: [Google]

50KB, 900x900px

>>60344618
>>60344618
FUCK YOU AMD FOR PUSHING THIS MORE CORES MEME. GOD FUCKING DAMN IT. I'M DONE WITH ALL THESE INFINITE CORES AND SHIT. WHY DON'T PUSH SINGLE CORE PERFORMANCE. YOU PIECE OF SHIT FUCKING DAMN

>>60345513
They'll be segmented by clockspeed, you'll probably have a cheapo 12 core at $600 or less with lower clocks.

>>60345544
Fuck off /v/

>>60345555
Quads of truth.

>>60345546
That's unattractive as well though, unless it can OC reliably, which isn't really what we've seen from Ryzen so far. I have no 12C workloads that I do frequently, getting something with low clocks wouldn't be very appealing.

>>60345546
>cheapo 12 core at $600 or less with lower clocks
SUPER ULTRA HYPER MEGA TASKING

>>60345584
Then why do you care about these chips? Get a quad core.

>>60345584
Too bad, because Intel's 12 cores will cost $1500 for 300 more MHz

>>60345601
I have a quad core, 6C or 8C would be useful to me, but 12C is likely going to push the clocks too low to be worth it unless the chip is amazing at OC with the proper cooling.

>>60345607
B-but more gihagurtz goyim. Buy our Genuine Intel processors.

File: [Mori] Kiniro Mosaic Pretty Days [720p].mkv_snapshot_12.34_[2017.03.05_14.42.08].jpg (114KB, 717x720px) Image search: [Google]

114KB, 717x720px

>>60345544
>mfw it's already beginning >>60330282

File: WP_20170507_11_57_09_Raw.jpg (2MB, 2014x1342px) Image search: [Google]

2MB, 2014x1342px

>>60344618
Confirmed.

>>60345622
These Ryzen chips are all the same die, all of them can OC to 4.0/4.1, overclocking 12 or even 16 cores to that would be insane performance.

Only issue is cooling.

>>60345644
>160W

Rofl

>>60345642
dumb karenshiiter

>>60345644
Intel becoming the new AMD

>>60345670
Is that 160? I can't read the blurry smeared text well enough.

>>60345656
16 cores at 3.9 would even be faster than the $7000 24 core Broadwell-E, which is some.. 2.3GHz?

Power usage aside.

File: o7hUlyV.png (186KB, 652x372px) Image search: [Google]

186KB, 652x372px

>>60345686

>>60345644
>4c/4t
>112W

Intel should die in a fire together with their clinically retarded engineers.

>>60345695
Read the OP

 - Dual DDR4-2666 for Kaby Lake-X / Quad DDR4-2666 for Skylake-X
- 112W for Kaby Lake-X / up to 160W for Skylake-X
- Apparently all Core i9 parts support AVX-512 (TBC)
- Launch in June, except 7920X (August)

>>60345644
This piece of shit company needs to be burnt to the ground

>>60345644
>4.3GHz all cores stock on 8C/10C chips
That's actually pretty impressive. The price is probably going to be ridiculous which will make them unappealing, a damn shame.

>>60345725
Yuo don't liek more ghz? You want longer bibelines?

>>60345701
In theory, 24core would win every time.

>>60345738
Chances are pretty high if they are doing overclocking tests of this garbage at their manufacturing plants.

>if you don't care about housefires, Intel has added Turbo Core 3.0 to our chips

>>60345751
In theory, but the clockspeed deficit is larger than the core deficit, and even in datacenter 50% more cores is not 50% more scaling.

High clocks aren't efficient though, so the 24 core clearly wins out at perf/watt

>>60345744
Isn't that 3.3 GHz base 4.3 GHz boost?
>>60345765
Will Coffee Lake-S get Turboboost 3.0? Or is that never showing up on consumer CPUs?

>>60345794
What exactly is turboboost 3.0? XFR: Intel edition?

>>60345744
All cores? Turbo boost has always been single core, all core will probably be +200MHz over base clock if that.

>>60345807
TB3.0 is single core only boost, it also doesn't give a shit about TDP so it's not enabled by default

As someone who has been on the intel train since the 920 days... intel are fucked. AMD actually did it the madmen.

>>60345812
>>60345794
>Isn't that 3.3 GHz base 4.3 GHz boost?
It's 3.something base, but it says "Maximum mit allen Kernen" under the Turbo 2.0 column, which translates to "maximum on all cores". The actual top turbo clock seems to be 4.5, which the CPU won't hit on all cores (at stock).

File: 281453CB00000578-3057819-image-a-43_1430188997605.jpg (119KB, 962x641px) Image search: [Google]

119KB, 962x641px

>>60345644
>>60345670
>>60345762
>>60345765
WHAT THE FUCK IS INTEL EVEN DOING

File: CPUStore8KFullHDDONOTSTEAL.png (255KB, 719x720px) Image search: [Google]

255KB, 719x720px

>i9
Intel gotta be THIS desperate right now to even name their cpus like this

>>60345319
If AMD's 16c32t series doesn't have 6 channel memory, it's DoA.

File: Screenshot from 12-05-2017 17:37:22.png (105KB, 692x704px) Image search: [Google]

105KB, 692x704px

>>60345849
I'll believe that when I see it, Intel had tried very hard to hide all core turbos for the last 5 years, but there are smarter people out there.

See this image.

>>60345890
Their FAVOURITE thing. Damage controlling with housefires.

File: goo.gif (1MB, 240x232px) Image search: [Google]

1MB, 240x232px

>>60345838
The Athlon days are back.

>>60345916
Skylake-X is quad channel, so is AMD's HEDT


Skylake Xeons are 6 channel, AMD's Opterons(tentative) are 8 channel
And these Opterons go from 16 to 32 cores.

>>60345929
Shitwrecker did it. Now we need poo in the loo king to deliver another Evergreen.

File: housefire inside.png (126KB, 357x330px) Image search: [Google]

126KB, 357x330px

>>60345644
>one hundred and sixty watts

File: 440.gif (2MB, 467x468px) Image search: [Google]

2MB, 467x468px

>>60345555

Quads of truth, [2]

>>60345910
irony part is that AMD HDET will probably be R9

>>60345938
Well, at least AMD is going to come out with something that might actually be able to outperform the 6950X.

Anybody know the prices for 16c32t from AMD? I might just sell my current rig for it.

File: 1494561530611.png (254KB, 484x471px) Image search: [Google]

254KB, 484x471px

Unbelievable, AMD may actually fucking do it

>>60345938
I haven't considered the possibility of 4 die MCM server chips at lower core count, that's gonna be some amazing density and DIMM for middle range core count chips.
Though I don't doubt there'll be 16 core chips that don't need so much memory and bandwith so they'll go 2 die MCM


This MCM thing is really fucking flexible.

>>60345995
Guessing it'll be anywhere from $1000 to $1200.

>>60345995
It's not gonna be cheap because AMD can allow to throw hefty price premium on it because Intel simply can't compete with it. Like Athlon FX back then.

>>60345918
I guess the image could be fake, but that's what it says. They don't have much reason to hide their clock if it's actually 4.3GHz on 8/10C, that's something fairly impressive which they'd want to show off.

>>60346019
And all it took was 5 years of Intel sitting on their ass while AMD made a new architecture that wasn't complete trash like Bulldozer was.

>>60345995
Realistically $1000-$1300
If they wanna go ultrajew over $1500

These are chips that can compare to Intel's 24 core Xeons due to sheer clockspeed, these Xeons cost over $5000, so they might command a premium

File: 4c3e9f87d222d4b818759ce7239c7d95[1].jpg (59KB, 640x480px) Image search: [Google]

59KB, 640x480px

If Zen+ improves the flaws of the current ryzen, AMD will fuck the jews in the ass.

File: 80OQlsm.png (239KB, 519x456px) Image search: [Google]

239KB, 519x456px

Stuttering housefires

>>60346062
It will, for fucks sake Papermaster himself it's gonna be tock-tock-tock for Zen.

>>60346105
Nice to know AMD won't be sitting on their ass, the real slaughter will start with 7nm though.

>>60345513
I would actually be okay with AMD 12c at around $800 or 16c at $1000. I could stretch the budget for that. What I'm not down for is paying more than $1000 for a 10-core chip.

>>60346134
>IBM's 7nm designed to run at literal HOUSEFIRE clocks for POWER
Ah yes, the clockspeed. The last bastion of Intel.

File: ufdiLMR.png (856KB, 801x1500px) Image search: [Google]

856KB, 801x1500px

NETBURST 2.0 CONFIRMED!

>>60346164
Is it gonna be SOI or Finfet?

>>60346190
>Is it gonna be SOI or Finfet?
SOI Finfets.

File: Screenshot from 2015-06-28 13-54-32.png (76KB, 405x329px) Image search: [Google]

76KB, 405x329px

>>60346200
Has science gone too far

>>60346211
It's fucking IBM. What did you expect, sanity?

>>60346164
To be fair, it's only going to be used in mainframes and the people bamboozled into buying POWER servers.

>>60346200
I don't think that's possible but it would be pretty insane.

>>60344618
>no prices listed
good goy

Wait, this isn't coffee lake, it's HEDT Skylake-x. Where is muh 6 cores on mainstream platforms (z370 or whatever)?

All they did was add i5 and i7 to their HEDT line and change the naming scheme. The only new processor here is the 7920x. Why is the 4c4t named i7? None of this makes sense.

>>60346232
But it's also a fact POWER is high clocking, and POWER10 will be on 7nm which GloFo bought off IBM, it's pretty much safe to conclude the 7nm Zens will be highly clocked.

File: rcj_Freescale_FinFET,-FD-SOI,-field-effect-transistor,-Silicon-on-Insulator.png (204KB, 460x325px) Image search: [Google]

204KB, 460x325px

>>60346240
Son...

>>60344862
add $500 to those prices goy

>>60346268
GloFo for 7nm due to IBM merging their foundry business with them.

>>60345642
>Thread deleted
>Mods confirmed Intel shills

>>60346272
Dear God. They did it? ABSOLUTE FUCKING MADMEN SOI FANATICS.

File: stm1.jpg (51KB, 886x492px) Image search: [Google]

51KB, 886x492px

>>60346272
It's known that SOI and Finfet are rather similar
There's even a joke they're the same shit just turned upside down, so I don't see how you can combine them.

>>60346288
It's archived, idiot. Deleted threads go straight to 404.

File: rip.jpg (74KB, 858x536px) Image search: [Google]

74KB, 858x536px

>>60345644

>>60345644
holy shit that L3 cache is a joke

>>60344924
Dude, the 205w part is probably half way to 4ghz. Clocks matter a lot more than you seem to think. You could easily double power consumption by going from 3ghz to 3.6ghz.

With 20 cores you can have more performance than the i7 6950x at less than half the power consumption, and that's exactly what Xeon processors do.

Consumers with muh gigahurtz get the shittiest power efficiency possible.

>>60346315
That's because they have 1MB L2 now.

File: jews.png (24KB, 596x643px) Image search: [Google]

24KB, 596x643px

>>60346274
My mistake.

>>60346357
good goy

>>60346329
>205w part is probably half way to 4ghz.

Eh, that's the 28 core Intel with 2.5GHz clock, at 205W TDP

>>60346342
Software can't fucking use 512kb L2 of Zen properly, why did they change it on Skylake-X?

>>60346342
>Implying that one can compensate for the other

>>60345531
I'm glad this still gets reposted ~That Polish guy here

>>60346403
Yikes, that's a triple strike, TDP, clocks and cores are inferior, and Naples clocks are based consumer B1 revision with no cherrypicked dies.

Getting into I/O and memory capacity and bandwidth just gets ugly.

>>60346436
Two reasons.

AVX512 will be thankful for it.
Big chunks of shared L3 are costly and yields are shit, POWER9 is the first one to abandon that design, followed by AMD Zen and finally Intel with Skylake-EP

>>60346513
>muh AVX512

>>60346533
Well, Intel is betting real hard on it against Teslas, they don't want them to eat their HPC lunch

>>60346513
Who the fuck uses meme512? Why would you EVER need vectors that big?

>>60346403
Dunno what Amd is doing then. Maybe fake tdp or clock figures. I know they have a slight advantage but it's not that big.

>>60346563
You don't fake figures to your enterprise clients unless you wanna get sued.

>>60346563
>I-If AMDs offer is better is b-because they're lying

>>60344618
DELID THIS

>>60346619
You can't, it's soldered.

>>60346583
Fake as in apples to oranges comparison. Measured differently, different load, etc. It's all explained in disclaimers. Might not be AMD who is fake either. My Intel Xeon has 105w tdp but only takes about 80w grinding avx, so that 105w clearly isn't correct.

File: 1488576947538.png (64KB, 960x540px) Image search: [Google]

64KB, 960x540px

>>60346563
Dude, have you seen the power figures of Zen under 3.0GHz?

It's a fucking monster.

>>60346627
soldered with scrap metal?

>>60346403
RIPtel

>>60346563
AMD measures their TDPs differently so expect a 100w AMD processor to be the equivalent of a 110 or 115w Intel processor.

>>60346631
>AMD hasn't released a 35W 8 core

What a fucking waste.
Why is AMD marketing so shit?

File: Screenshot from 05-03-2017 23:38:48.png (62KB, 957x532px) Image search: [Google]

62KB, 957x532px

>>60346630

>>60346631
No I haven't. Looks pretty good. 1400 is less than 80w with 8 cores. I need 12 for it at that tdp.

>>60346680
You mean threads

>>60346659
People would cry about it being bad for gaymen

>>60346699
No cores, both ryzen and Xeon have Ht and 1800x is 8 core.

>>60346705
Market it for low power workstations.

>>60345644
>new socket

HAHHAHAHAHAHAHHAHAHAHAHAHAHA

File: 1445873750669.jpg (15KB, 250x250px) Image search: [Google]

15KB, 250x250px

>>60346726

>PGA
>2017
when will these poojets learn

>>60346721
mUH GAEYMS

>>60345644
It's bibeline all over again

File: 1227-2.jpg (191KB, 750x500px) Image search: [Google]

191KB, 750x500px

>>60346757
>implying Naples and HEDT Ryzens aren't LGA

>>60346785
That's a big socket.

>>60346785
Naples and HEDT ryzens need to be LGA, they have way too much pins to be safe putting them on the processor itself

>>60346801
For you

>>60346801
It houses literally 806mm2 of silicon, with a shitload of memory channels and I/O, of course it is.
The cores only need around 60% of those 4000 pins, the rest is I/O

>>60346630
Intel uses TDP as the maximum draw. You'll probably never find a workload that can actually hit that. AMD's TDP is the equivalent of Intel's SDP, which is actually the industry standard definition of TDP. Intel's way is actually useful though, while the other way is just a relative performance indicator.

>>60346801
Intel's Intel LGA 3647 or whatever number is also pretty huge too.

>>60346717
1400 has four physical cores, not eight

File: Screenshot from 12-05-2017 18:34:43.png (76KB, 1220x884px) Image search: [Google]

76KB, 1220x884px

>>60346848
Eh...
From ark.intel

>>60346851
1400 points in cinebench, not the r5 1400

>>60346851
He was talking about the score
1400

>>60345644
Presshot all over

>>60346864
HAHAHAHAHAHAHAHAAHAHAHA

>>60346864
lol

>>60346875
>>60346878
well, shit, if i ever meet you guys you're getting a free blowjob as compensation

>>60346848
There's no standard for TDP, it's literally a figure that changes every generation whenever someone farts.

You just have to look at the 15W mobile chips.

>>60346801
4U Rackmount

>>60346659
Because you can buy any R7 and run it at 35W cTDP

File: 1488432385535.png (655KB, 1920x1200px) Image search: [Google]

655KB, 1920x1200px

>>60346631

>>60346939
There's no reason these can't fit in 2U, in fact the test system for the Naples presentation was done in a 2U

>>60346954
Are you retarded or merely pretending?

>>60346972
I have no reason to go along with your bane memes

>>60344618
This looks good, the issue is that this will be a platform that's gonna be 2-2.5 years and it seems to still come short of the Zen's efficiency.

The bigger issue is that AMD plans to release new iterations of Zen every YEAR, and a simple 10% IPC increase with Zen+ next year would put Intel's entire Xeon lineup in the grave.

>>60346547
Some extremely niche cases of HPC, probably not nearly enough to sustain a feature, much less a product
>inb4 Xeon Phi's
They're somewhat popular only due to x86

>>60346547
It's Intel's desperate attempt thinking they can use x86 to match the sheer throughput of GPUs

File: THE_BOMB.jpg (80KB, 635x636px) Image search: [Google]

80KB, 635x636px

>>60345644
Intel finally switched places with AMD
What a time to be alive

>>60347200
hownew.ru

>>60344741
>>60344656
/thread.

Its just abortion Xeon chips.

File: C5qPYNbUoAAVYlc.jpg (111KB, 1000x1000px) Image search: [Google]

111KB, 1000x1000px

>>60344618
>he doesn't list a price or power consumption

>>60349745
Bad goy

>>60349745
There's an image with that shit listed and it's also in his description

File: megatasking_intensifies.jpg (72KB, 493x493px) Image search: [Google]

72KB, 493x493px

FUCK YEAH MEGA-TASKING

How much will the 7800X and 8700k cost?

>>60351254
1. A lot
2. More than a 7700K

>>60351254
Less than a 6800K and 7700K, respectively

File: 1461688072353.gif (746KB, 500x281px) Image search: [Google]

746KB, 500x281px

>>60351314
>Intel
>Lowering prices

>>60351337
>this kills the AMDrone

>>60351353
>This kills the delusional Intelcuck that thinks they'll finally drop their prices this time, for real we swear

>>60351417
They have to, now there is competition.

The situation is completely different from the past 6 years.

I wouldn't expect a simpleton to understand though

>>60344618
>tfw zen is actually more efficient
>tfw these will probably run hotter than the zen parts
I won't fall for the wait meme once more

>>60351451
If they haven't dropped prices already, they never will.

>Intel is allegedly preparing six HEDT processors, two based on KabylakeX architecture and four based on SkylakeX. The most powerful CPU called i9-7920X will launch in August, while the rest will be available in June. The processors will have TDP up to 112W (KabylakeX) and 160W (SkylakeX)

AMD on suicide watch

>>60351475
>160W
>205W
you bet your ass they will

>>60352178
but outside of demanding tasks wattages will be in single to low double digits, does it really matter what max draw is for most people?

>>60352201
you don't buy a 10-core chip to have it idle at the desktop for the whole day, do you?

File: Intel-Core-i9-7920X-Vs-Ryzen-7-1800X_1-1140x854.png (220KB, 1140x854px) Image search: [Google]

220KB, 1140x854px

i9 12 core at 2.9ghz demolishes 1800X at 4.1ghz

KEK

guess what the i9 8c/12t does to AMD at 4.5ghz

>>60352361
*8c/16t

lmao AMD is so done

File: AMD-Zen-Feature-watermarked-1920x1080.jpg (123KB, 1920x1080px) Image search: [Google]

123KB, 1920x1080px

>>60344618
>Core i9-7820X
>turbo boost 2
2 cores @ 4.3ghz
>turbo boost 3
one core @ 4.5ghz
>all core usage
3.6ghz
this crushes ryzen r7's, unless, its priced at typical jewtel prices. most likely priced ~$550 - $650 which will make the 1800x irreverent.

>Core i9-7800X
>turbo boost 2
2 cores @ 4.0ghz
>all core usage
3.5ghz
this will be a strong contender to the ryzen r7's. most likely will be priced ~$400 - $450. the r7 1700 non-x will be a better value outside of gaming in this price bracket.

the 10 core will most likely take the $1,000 slot with the new 12 core, 7920X, taking the $1,500 slot.

the quad cores will be priced at $250 - $350 respectfully. intel won't lower their price because they are "faster" in single threaded performance against the 7800x.

so my predictions....
>Core i7-7640K
$250
>Core i7-7740K
$350
>Core i9-7800X
$450
>Core i9-7820X
$550 - $650
>Core i9-7900X
$1,000
>Core i9-7920X
$1,500 - $1,700

1800x & 1700x irreverent
1700 safe but will be the budget multi-threaded processor
r6's and r7's will pretty much stay the same

intel's own kaby lake mainstream platform absolutely BTFO and finished. completely 100% irrelevant.

>pic related
a path of zen leading to the landfill

pretty impressed. intel will be coming out hitting hard.

>>60352361
There's something seriously screwy with both these results. There's not a snowball's chance in hell Intel just suddenly increased their IPC by that much that a 3 GHz turbo can beat 4 GHz turbo on the 1800X.

>>60352201
are you fucking implying that the 7700k maximum power draw is 91W you delusional nigger?

>>60352539

>sauce

>>60352539
you know actually i really believe the 6 core 7800X will probably be faster in multi-threaded against the 1800x when overclocked to 4-4.1ghz.

really the entire r7 line just got BTFO and finished by skylake-x.

all amd will have going for them will be the low end budget line. pretty sad but at least they have these few months.... but then again they shouldn't. we should make sure to get the word out that intel will be releasing a 6 core that's faster in every single scenario than ryzen.

i can't stomach people buying ryzen knowing it will be money thrown in the trash by this.

>>60352659
common sense.the 12 core part is to replace the current 10 core part. intel isn't removing a part. unless they want to charge 2k for that 12 core part.

but seeing how amd released something at the moment that is competitive and is hurting their 8 core, $1,000 part sells, they will be reducing price. this is also helped proved by the notion intel is chopping off pci-express lanes. the 8 core is even limited to 28. this is to save money to increase yields.

sooo...
>Core i7-7640K
$250
>Core i7-7740K
$350
>Core i9-7800X
$450
>Core i9-7820X
$550 - $650
>Core i9-7900X
$1,000
>Core i9-7920X
$1,500 - $1,700

the 7820X will completely destroy all ryzen r7's. the 6 core 7800X will be a brutal competitor to ryzen r7's. based on skylake, that's a solid ~7% increase in single threaded performance over broadwell-e but higher clocks. a overclocked 7800x at 4-4.2ghz will be on the heels of ryzen r7. the 1800x will most likely come out ahead but its more money... and its in the 7820x territory.

ryzen really DOA.

>>60352361
Yeah these benchmarks seem very much like a big pile of bullshit.

http://cpu.userbenchmark.com/Compare/Intel-Skylake-X-10-Core-HEDT-10C-vs-AMD-Ryzen-7-1800X/m233971vs3916

http://cpu.userbenchmark.com/Compare/Intel-Skylake-X-12-Core-HEDT-12C-vs-AMD-Ryzen-7-1800X/m278103vs3916

The 10 core version has the same single core as the 1800X at 3.1 GHz no turbo, then the 12 core version has significantly more single core at 3.0 GHz with turbo. There's something fucky going on here.

>>60352665
>the common end user AMD's platform got destroyed destroyed by Intel's HEDT
do you realize that they are completely different products for different consumers?
Intel's HEDT is going up against AMD's HEDT, not the final consumer r7 line

>>60352751
>>Core i7-7640K
>$250
>>Core i7-7740K
>$350
>>Core i9-7800X
>$450
>>Core i9-7820X
>$550 - $650
>>Core i9-7900X
>$1,000
>>Core i9-7920X
>$1,500 - $1,700

lol if you really think these prices are anywhere near accurate. It's their HEDT platform, you better add at least $250 to all of those, even the shitty i5 and i7 rehash.

>>60352775
price. these parts are filling in current roles. intel is releasing quads. a 4 c / 4 t.
>Core i7-7640K
$250
>Core i7-7740K
$350
>Core i9-7800X
$450
>Core i9-7820X
$550 - $650
>Core i9-7900X
$1,000
>Core i9-7920X
$1,500 - $1,700

ryzen has been demoted to sub $200 trash. deal with it.

File: intel_xeon_scalable_press_deck_lp_embargo_5.4.17_at_5am_pdt-page-013_575px.jpg (131KB, 955x537px) Image search: [Google]

131KB, 955x537px

>>60352361
Yeah, not buying it.
Even their own marketing slides put it at 30% better than lower clocked Haswell Xeons.

>>60352827
>AMD's HEDT and Naples doesn't exist
this is you

>>60352813
you do know these prices are pretty much their current HEDT platform right?
>Core i7 6800k
$450
>Core i7 6850
$550
>Core i7 6900x
$1,000
>Core i7-6950x
$1,700

fill in the wholes with replacements:
>Core i7-7640K
$250
>Core i7-7740K
$350
those two above are NEW the HEDT platform

these guys are the traditional.
>Core i9-7800X
$450
>Core i9-7820X
$550 - $650
>Core i9-7900X
$1,000
>Core i9-7920X
$1,500 - $1,700

but seeing how amd released something at the moment that is competitive and is hurting their 8 core, $1,000 part sells, they will be reducing price. this is also helped proved by the notion intel is chopping off pci-express lanes. the 8 core is even limited to 28. this is to save money to increase yields.

ryzen is actually truly BTFO and finished. demoted to sub $200 trash.

>>60344713
X99 supported 8 PCIe 2.0 links from the chipset, plus the PCIe from the CPU.

X299 supports 22 PCIe 3.0 lanes from the chipset PLUS the PCIe lanes from the CPU.

So 22 + 28, or 50 PCIe lanes for a base configuration.

22+44 for 66 PCIe lanes using the 10 core.

File: Screenshot_20170512_235309.png (120KB, 1193x865px) Image search: [Google]

120KB, 1193x865px

>>60352754
Yeah. This looks legit.

>>60352566
>>60352754
>>60352861
>>60352916

>that damage control by AMD drones

holy shit

You will enough why Intel comes out hitting hard in June when multiple tests are done

>>60352888
The fact that AMD's consumer R7 CPUs even compare to Intel's overpriced HEDT shows how badly they're hurting. And AMD's HEDT platform is going to have 16c/32t. And more PCIe lanes.

Also, those benchmarks you posted are bullshit, so don't go running around trying to claim AMD is about to get BTFO by the most unforseen IPC increase ever.

>>60345644

i9 housefire time

File: 1492893633027.png (158KB, 2033x909px) Image search: [Google]

158KB, 2033x909px

>>60353015

>>60345644

all cores at 4.5 ghz

sounds juicy to me time to upgrade

>>60353015
look, i'm rolling a 6600K here, but you're delusional if you think intel's current architecture isn't nearing its' end
it's literally P3 that was improved over time and there's just so much they can do before they reach limitations, like what we're seeing now
112W 4 core processors are disgusting compared to what AMD came out with

>>60352887
>AMD HEDT
irreverent as intel's HEDT platform will be priced aggressively to shit all over current ryzen. amd will have to replace current ryzen with their HEDT platform to compete.

who in the hell would buy a 1800x when the 7820k will shit all over it for nearly the same price? the 1700x is pure trash. the 1700 will still fit a niche role. will depend on how well the 7800k will overclock. seeing those boosts though it most likely will be able to hit 4ghz minimum making the 1700 really a poor mans multi-threaded cpu.

only the r6's and r5's will fit a role and continue to sell well. unless coffee lake new mainstream utterly BTFO of them at the $250 and less price brackets. which they probably will if intel is able to up the clocks. which it seems like intel will be able to do.

really ryzen will be shat on hard in august. zen+ better bring ryzen up to kaby lake level of performance in single threaded.

>>60353044

?

>>60353074
That is not how turboboost works.

>>60353096
Keep pondering a little more.

>kikes are defending housefires now
Lmao look at this >>60353087

>>60353107

You are mad it's okay. You will be crying in June.

>>60353025
i never posted benchmarks and AMD did get BTFO and finished with skylake-x, x299. intel is a sleeping giant. you fools should know this.

intel got bitch slapped and came out firing nukes with the core 2 series. then the i series. you back intel into a corner intel can do some pretty surprising things.

this case intel can't squeeze out extra performance, so they are going the core count and reduce pricing. intel will not let go any marketshare. amd better enjoy their short lived sucess. but i will make sure to spread around forums, reddit, and here how amd will be finished by august and best not to waste any ounce of money right now. better to wait for you can get a 6 core thats faster even in multi-threaded than the 1700x.

>>60353131
Don't worry I'm sure Intel will gift you a 20% IPC increase.

File: c077f42a6b0c28106928204ce65943049d77a9e96c997b5f39df7b06a8ea82d5.jpg (96KB, 500x649px) Image search: [Google]

96KB, 500x649px

>>60353131
lmao cuck'd by housefires

>>60353099

what did he mean by this? I can't OC my cores to 4.5ghz?

>>60353137
It's sleeping housefire.

>>60353171
Maybe if you're ready to burn your house down.

thank you Intel

Unlike cancer like AMD is with their 180W 8 cores you deliver once again 12 cores at 160W

Perfect and shitting on them in every way.

File: 1315867163855.jpg (33KB, 500x412px) Image search: [Google]

33KB, 500x412px

>13mb of L3 cache
The troll does realize the broadwell X chips currently have 25mb of L3 right?
6950x

>>60353173
its intel. no one cares about house fires when it comes to nvidia and intel. they could truly burn down your entire neighborhood and people will still bend over backwards.

suck it amd fags. you guys are BTFO.

>>60353171
you can
one of them

>>60353173

>le housefire
lost count of houses you burned down your your ATI cards and Athlons ?
kek

>>60353214
thats for turbo boost, not OC on all cores Rajeet

File: reach for the bleach.jpg (12KB, 256x256px) Image search: [Google]

12KB, 256x256px

>>60352361
>>60352382
>>60352539
>>60352665
>>60352751
>>60352827
>>60352888
>>60353015
>>60353087
>>60353131
lol I have never seen one Intelcuck spam so much nonsense in such a short amount of time. I was waiting for the damage control squad to show up, here they are.

File: single socket holocaust.jpg (28KB, 550x550px) Image search: [Google]

28KB, 550x550px

>>60353245
Ha ha well memed my friend here have a CPU

>>60353203
>Unlike cancer like AMD is with their 180W 8 cores you deliver once again 12 cores at 160W
Their 16 core is running at 180w.

>>60352754
What's with this magical 12 core that's 20% faster in single core at the same clock than the 10 core of the same arch?

>>60353263
i was waiting for fool amd fags who thought intel couldn't release a $550 8 core.

you poke a sleeping giant, you're going to awake it. thats intel.

with vega getting bitched slapped by volta, and now ryzen anally raped by skylake-x, amd will be BTFO and finished.

>>60353265
>he goes all out with Intel pictures
mad as fuck
Intel domination

>>60353310
(You) are not even trying anymore.

>>60353263

I'm not the one damage controlling the 1800X you cuck

>>60353307
That 7820X is going to cost at least $750. Who are you kidding?

File: 1385420441583.jpg (251KB, 852x939px) Image search: [Google]

251KB, 852x939px

>>60353310
Intel is going to get BTFO by amd on the server side

>>60344862
>shitty prices
>just playing catchup to amd's multithreaded performance
>housefires
no thanks im going to get 7nm ryzen to get rid of my 4790k

>>60353325

speak for yourself autismo

>>60353044
Not a valid bencfhmark

>>60353297
Don't know, it's some fake shit someone uploaded for lulz or whatever.

>>60353087
Why are you repeatedly comparing consumer parts to HEDT parts? You're not considering platform costs at all, nor are you considering the current cost of Intel's HEDT product stack; a 6900k at nearly $1100 is on par with an 1800x at less than half the price. You're trying to say that Intel is going to double the value it can offer to consumers without moving to a new process node. That's complete lunacy. Over the course of the past six years, Intel has barely been able to increase the value to consumers by 40%. Although you're trying to make it seem like you have a reasoned approach to this, it just isn't so.

And by the way, if you somehow manage to figure out a way to worm your way through some sort of argument for your idiotic pricing model, can you please explain to me why Intel would gut its own Xeon lineups by undercutting parts like the E5-2687W so heavily? Are you going to also make the case that Intel can offer twice the value to the server market?

>>60353371
No it's not, it might just be detecting clocks wrong.

>>60353368
And I'm sure the other ones posted ITT are, right?

File: Screenshot from 2017-05-08 22-57-34.png (92KB, 883x363px) Image search: [Google]

92KB, 883x363px

>>60344618
>180 W TDP

>>60353388
Even if you go by what the leaked clockspeeds supposedly are, it still makes no sense.

File: 1490404105649.png (944KB, 1228x1502px) Image search: [Google]

944KB, 1228x1502px

>>60353087
keep dreaming pajeet

>>60353413
it's 160w Rajeet still less than AMD 180W 8 cores under load

most likely take even less in standby

>>60353392
Not him, the 10 core userbenchmark results looks fairly normal, the 12 core doesn't make sense as it's a shitload faster at same clocks.

>>60353449
see
>>60353286

>>60353449
>AMD 180W 8 cores under load
>lying on the internet
wew

File: 1491143613039.png (1MB, 1110x5110px) Image search: [Google]

1MB, 1110x5110px

>>60353449

UUUUUUUUUUUUUH INTEL

i9 7820 looks JUICY AS FUCK

THAKN YOU INTEL THANKS

File: ryzen power.jpg (54KB, 711x533px) Image search: [Google]

54KB, 711x533px

>>60353449
OwO

>>60353506
If Tom's useful for anything it's getting direct power measurements from the rails instead of inaccurate software sensor and wall readers most "reviewers" use

>>60353451
Intel and AMD's IPC are pretty much dead even at the moment. It would require the Intel CPU to be clocked at 3.6 GHz, but the leaked clocks say 3.3 GHz, and the benchmark says 3.1 GHz. That's not even counting XFR on the 1800X pushing clocks even higher, up to 4.1 GHz. It doesn't appear turboboost was enabled for this test, but if it was, using the leaked turboboost would put it at 4.5 GHz turbo, which means it's losing to the AMD CPU.

>>60353539
>auto cad 2015
>medium workload

hahahah

is this Rajeet even trying

Don't make me look up those Prime numbers

Intel wins every time

>>60353605
>says this, while ignoring the prime95 results 2 posts above

Here's your reply.

>>60353506
>OC'd intel takes as much as 1700 stock

lmfao

>>60353631
see>>60353635

talk shit later Rajeet

>>60352167
>TDP up to 112W (KabylakeX) and 160W (SkylakeX)
Intel rates tdp at base clocks, not boost clocks. You better hope your fire suppression system is finished first.

>>60353656
Don't make shit up, it's quite disgusting.
They literally had to undervolt the 4 core 7700k to compete on power consumption.
A fucking quad core.

File: power_multi_thread[1].png (11KB, 500x250px) Image search: [Google]

11KB, 500x250px

>>60353605
this is what you were meant to post
>8c/16t cpu using 2W more than 4c/8t cpu that's slower
holy fucking shit intel get it together

>>60353605
OwÓ

>he fell for Ryzen

sad really

>>60353697

>4.0 ghz
193 Watts

my fucking sides in orbit

>>60353697
>1800X
>4 ghz
>200w

lmao

>>60353768
go ahead, post the TDP of a 4.0 6950X and we'll all laugh

>>60352861
Why aren't they comparing it to Broadwell but Haswell?

File: lmaoooooo.png (306KB, 944x533px) Image search: [Google]

306KB, 944x533px

>>60344618
>Core i9
The ultimate housefire. The promise of nuclear fusion at home has finally come true. Tejas LIVES.

>>60353804
Because 15% looks less impressive than 30%

>>60353789

Skylake X: 160w 12 cores
AMD cancer Ryzen: 8 cores 200w

Hmm

>>60353768
>>60353786
Your parents need to take away your computer, desu.

>>60353786
what's weird about that?
8C chips are bound to draw a lot of power at those clocks
besides, this is AMD's first tackle at modern chips and it'll get better with newer revisions like Zen+, compared to Intel's architecture with its' roots back in P3 days

i honestly regret buying a 6600K and Z170, wish i had waited a year to get Zen

File: OH MAN OH FUCK.jpg (21KB, 304x294px) Image search: [Google]

21KB, 304x294px

>>60353822
>160W for Skylake-X
>>160W for Skylake-X
>>>160W for Skylake-X
>>>>>160W for Skylake-X

>>60353829
>Skylake X: 160w 12 cores
how do you explain 112W 4-cores then?

>>60353873

better performance higher clocks
(which AMD has none of)

You people do know you have a report button, right?

>>60352908
>X299 supports 22 PCIe 3.0 lanes from the chipset
Yeah, but it's all fed through a 4 lanes of 3.0, so it doesn't matter for shit how many lanes Intel cram on the chipset.

>>60353907
>better performance higher clocks
>in the age of parallelized workloads
that's no excuse for a housefire larger than 9590

>>60353829
>TDP is power consumption
Are we really going to play this game again?

>>60353844
I don't quite frankly do not care what you regret or not
6600k existed before AMD shit out anything

>>60353969
who are you quoting ?

>>60353975
that's exactly what i mean, i should've waited instead of buying a gimped 4c/4t chip
i'm satisfied with it at the moment, but it'll last me way shorter than a ryzen chip would
this is coming from someone whose upgrade path was
>p3
>athlon X2
>C2D E6750
>i5 6600K
not taking stuff like 486DX into account since i just used those computers and they weren't bought with my own money

>>60353822
CoffeeLake 50 stage BIBELAINNEN CONFIREMD

>>60353822
>>60353854
NOOOOO DONT TALK ABOUT THIS
STOP THIS

https://communities.intel.com/thread/26542

Holy fuck the retards, trolls and shills flooded the crap out of this thread.

>>60354128
It's one guy.

File: 1465590479231.jpg (81KB, 419x480px) Image search: [Google]

81KB, 419x480px

>all those butt blasted ryzen owners ITT

holy fuck lmao

>>60354198
3570k here.

After Madelline, you're possibly the most mentally deficient person to grace this board.
You're our very own clown.

damn Intel thanks

44 lanes
2666 ddr4 base
up to 12 cores

AMD still sells like trash, stock crashed hard

Nvidia + Intel combo only

>>60353697
>quad core with the same power consumption as a 8 core
AHAHAHAHAHHA

>>60352908
>X299 supports 22 PCIe 3.0 lanes from the chipset
And they're useless if you actually want to use them for something bandwidth-intensive, since the chipset only has 4x 3.0 worth to the actual CPU.

File: unspecified.png (24KB, 382x379px) Image search: [Google]

24KB, 382x379px

>>60354310
>tfw jewish and you read posts like this

>>60353829
>pootel shill going full damage control

>>60353907
>muh clock
reminds me of pentium 4

>>60354310
>44 lanes
On the $1000+ CPUs, they're actually jewing harder than with the previous HEDT lines.

>>60354304

kek
the only clown is your smelly mother ITT which is getting touched by a group of Rajeets

Good riddance my fellow Intel user.

>>60344713
AMD Whitehaven will save us, as Ryzen did.

It's honestly ridiculous how retarded all the intel shit is by comparison.

if Ryzen had Thunderbolt 3 I'd recommend it for absolutely everything

Reminder to report shitposters.

>>60354337
>5ghz
>152w
>better perf overall

vs

>200w
>8 core
>shittier perf

i'm dying over here hahaha