New socket for 14nm++++

X570 BAD
BUY OUR NEW MOTHERBOARD GOY

That five-digit model number though, it's obviously going to kick asses so you better pony up for a new mobo.

Attached: debiru side.png (309x286, 71K)

no one will ever believe those prices or the speccs
take a hike.

>don't release new products
and this why you don't make the big bucks nor work for a multibillion corporation like Intel

A socket change a year keeps the goyim in fear

There's no 10nm desktop parts, only for mobile

Intel will be using 14nm++++++++++ until they get their own 7nm parts right...or if their GPU launch with Samsung is successful they might pay them off or pay them to make their desktop parts while keeping mobile and Xeons in-house

Source?

Samshit gonna swim in money like never before with deals they made
Nvidia would be lucrative but now also Intel going for their 7nm EUV fabs? holy crap get some Samshit stock RIGHT FUCKING NOW.

>another 14nm++++++++++ rehash with the same hardware exploits/faults
>another "new" socket required
Intel cuc-err-customers sure do like getting milked out of their cash.

99% sure it was r*ddit hallucination, saw that kind of post before here.
thats NOT a leak until OP deliver a source

Intel goys are used to it.

Attached: 1469846418842.jpg (367x295, 18K)

>Spending millions on products that will never recoup their cost during their product lifetime is totally a smart business decision.

>Implying that Intel's senior management isn't still stuck with the mindset of the halcyon days during the Core-2 through Sandy Bridge domination era.

>implying that shareholders aren't going to do a around round of the guillotine in 2019-2020

>14nm+++
>10c/20t
>65W TDP
Kek, good luck with that intel

how much hertz will the i9-99999K have?

It's probably better to go with a new one than deal with people crying about why it won't fit in their boards despite being same socket.

Attached: intel galaxy xe.jpg (720x960, 241K)

>They are making late-era Bulldozer/Piledriver look sane by comparison.
That's some hardcore AMDroning here, with all shit intel has going on, they haven't even come near to Bulldozer disaster level.

they could literally rename their 14nm+++ to "7"nm and it wouldn't be further from real 7nm than TSMC's "7"nm is

Is 10900KF means kelvins or fahrenheit?

Why can't my s9 zoom in on photos anymore?

>doesn;t have game cache

Kentucky Fried

bout tree fiddy

i got a 7600k for $150 from fry's but honestly if i didn't buy that i would still be using a dual core pentium and would probably have just given up on going intel

i only did it because my old phenom motherboard died when the phenom 2 was still the highest performing amd chip they sold so there was really only one thing to do

Kek

>14++++++++++++
>hyberbreaded 10 gores
I really hope this is fake, else I'd fear for the poor souls that couldn't control the eldritch abominations they'd have to summon to cool it

>14+++ nm
fuck off intel

TDP DOESN'T MATTER!

Attached: 7f53425cbf.jpg (310x480, 74K)

my 7600k is overclocked to 4.5 with no consideration for core voltage and still on the dinky extruded stock cooler and doesn't go above 75c, i don't think there's going to be cooling issues with 10 cores

lmao

It's like they chose those TDP numbers just to match AMD numbers and now they'll have to go through mental gymnastic to redefine what TDP is so that the number actually means something.

I bet it's gonna be something like "idle TDP" or "single core load TDP"

That's probably just base clock TDP. I mean it's 2.7GHz. 65W is totally possible with 2.7.

Wonder if they have security fixes built into the actual hardware.

oy vey, new sockets are good for you goyim

>4.4 boost
It's going to suck.

That's more than a 3600 can OC lol

>march 7
obvious fake.

Attached: .png (650x182, 22K)

>i3's are now 4/8

NO 7700k BROS, WE GOT ROBBED

Zen 2 has much higher IPC so Intel needs higher frequency to match it.

>tfw everyone who bought a 9400F's and 9600K's just to spite AMD are all cucked out of HT.

Attached: tenor.gif (498x280, 1.19M)

>much higher IPC
>wins a few % in some tests, loses a few % in other tests
>MUCH HIGHER IPC
>M U C H
Kek, drones never disappoint

>wins a few % in some tests, loses a few % in other tests
Against 5GHz Intels, yes.

>Zen 2 has much higher IPC
no, Zen IPC is slightly less. Zen is probably winning at normalized frequency due to more cache. If Intel doubled its L2 and increased the L3 by 33% it would pull ahead at the same frequencies.

what point of Instructions PER CLOCK don't you understand?

Is a 9900K that stuck up your ass?

>(((((65w))))) 10c/20t
yea Right

Attached: (((95w))).png (1112x833, 74K)

>New socket for 14nm++++
AHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAH

Attached: 1491442602020.jpg (1024x576, 130K)

>what point of Instructions PER CLOCK don't you understand?

IPC is not a function cache misses.

Do you go the cellar? When can I catch you to see your full set?

yeah I saw that... crazy there is no way that's 65W unless they only pick best one silicion to be used out of all millions made

>moving goalposts

why am I not surprised.

You know the whole concept of measuring performance by IPC is retarded because you can technically have more IPC and still be significantly slower? For example, you can do the same job using regular x86-64 instructions or dedicated SIMD instructions if they apply. You'll run more of the x86 instructions per clock but they will be doing less actual work.

next thing you'll tell me is that the Pentium 4 is better than an Athlon X2 simply because Megahertz.

Uhh you misunderstood me to be defending intel, while all I'm doing is point out how retarded taking "instructions per clock" in literal sense is

ITS AFRAID

>SAME SOCKET
IMAO

You know, if someone goes Ryzen only now they may be just as screwed future-wise, because at this point it's now known whether there is still any future upgrade path for AM4.

AM4 probably has only one more generation

That's not even certain, AMD promised AM4 support until 2020 but never confirmed that includes 2020 CPUs

If it ever gets a 4th gen on AM4 i'm jumping to it. Should be a significant upgrade from my 2700X

South Korea and Japan are in full trade war now with Japan blocking sales of etching gas Samsung uses because S. Korean supreme court ruled the wording of 1965 Japan-S. Korea agreement on colonial reparation doesn't prevent individual victims who were sent to Japan to work in forced labor camps from suing the beneficiaries of their forced labor (i.e. Mitsubishi Heavy Industries). Japan has been REEEEEing really hard since the verdict, but they should have hired actually competent lawyers to draw up the agreement in 1965 instead of leaving a huge hole.

This is the same company that has historically been nice to people who want to upgrade with FM2,FM2+, AM2, AM3, AM3+, and now AM4

Well.. 6 and 7 series mobos were compatible. 8 and 9 ones too. Same goes for 2nd and 3rd. 5 gen was only compatible with the refreshed Haswell 9x series tho.
So there's usually 2 generation compatibility. Icelake support seems reasonable.

Yeah they've been so nice with AM3+, in term of both actual socket compatibility as well as it being an "upgrade"...

4th was only compatible with itself tho.
If you get 2 generations consider youself lucky.

You know LGA1151 compatibility bullshit stems mostly from the fact that intel was caught with its pants down and wasn't planning to significantly increase core count in near future. They just haven't planned to do that on 1151.

It's possible LGA1159 was always somewhere on the internal roadmaps, and the platform just wasn't ready at the time AMD raped them.

Intel has literally nothing up their sleeve except moar hertz and moar cores

Attached: 1562545301216.png (822x812, 71K)

When was the last time AMD surpassed Intel in IPC? I don't even remember, 2006?

inb4 99 vs 102

you can't tell the difference irl

Attached: k.png (1415x755, 724K)

Probably around 2005 before the Core 2 architecture was released.

no problem, just run it at 1GHz

But Zen 2 was MUCH better just a couple posts ago WHAT HAPPENED DRONE? MOVING GOALPOSTS AGAIN?

do you even understand what IPC means you complete retard

Yeah, about that...

Attached: AMD-Ryzen-9-3900X-and-Ryzen-7-3700X-CPU-Review_Assassins-Creed-Odyssey-713x740.png (713x740, 128K)

Intel fanboys are pozzed in the head so probably not.

What is the manufacturing cost of enabling HT on a CPU to Intel? Aside from higher quality wafer, what makes 8700k more expensive than a 8600k?

Attached: 1541240798399.jpg (408x439, 50K)

Here me out for a second but 65W isn't unreasonable even for a 10 core assuming TDP is strictly at the 2.7Ghz base clock
Intel has quad cores at 15W TDP @ 2Ghz base so even if you had two and a half of those that is still only 45W which leaves you 20W to add that +700Mhz to the base clock of the desktop part
May not be entirely how it works but makes sense in my head

>14nm++++
So that's 18nm then?

then they better get at least an 8 core. my 2500k lasted 7 years, an 8/16 cpu should manage at least that

16nm if we assume a "++" means "+1"

The guys @ Intel are next level trolls. You gotta respect that.

lmao

samefagging esl, guessing pinoy

If this is real amd is fucked, theyd have to drop prices pretty hard. The 350k would become the budget gayming king, not to mention intels entry level 6 core matches the 3600 and has an igpu for $20 less. Its also pretty epic how anyone who bought an intel chip in the last 3 years essentially paid 2x what its worth now.

Real question though is that if this lineup is real, why is intel letting amd steal the show with ryzen 3000? The 3900x is selling like hotcakes and intels 10core still might not even match it in multicore performance. Factor in the fact that you have to get a new motherboard anyway for these new intel chips, more people are going to keep jumping ship until this shit is officially annnounced.

It isn't real, it's a shopped slide
See

It's a microprocessor microcode update away. I think they just bin the worse portions of their dies and then turn it off in the microcode for the non K parts. Because if AMD can enable it through their entire lineup, there is no reason Intel can't. The only problem is that HT is now a security risk and needs to be reworked before it gets put back in a processor.

Bullshit. If you buy a k processor you are doing it to overclock and silicon Lottery says it will do 5ghz without issue. I wanted to buy a 3600 but you can't get anything more out of it AND you need to buy 3600mhz ram.

>What is the manufacturing cost of enabling HT on a CPU to Intel? Aside from higher quality wafer, what makes 8700k more expensive than a 8600k?
Nothing, it's artificial scarcity. Even clock speed is fake, odds are 99.9% of chips can run at the speed the top model does on desktop voltages. Why do you think there isn't 10 different versions of each iphone or android device with different clockspeeds? Unless you're running LN2 subzero cooling you will melt the CPU before binning becomes an issue.

oh totally

fucking retard

Its literally just segmentation, they have to go out of their way to disable it.

>I think they just bin the worse portions of their dies and then turn it off in the microcode for the non K parts.
I doubt they even waste the time to bin these things beyond checking they work, X86 CPUs are practically the only semiconductor product scummy enough to charge for mhz.