Intel's 8th Generation Core Family - Coffee Lake (LGA 1151, 6C/12T)

Where do you expect Core i7-8700K's Turbo to land?

  • 3.8/3.9 GHz

    Votes: 0 0.0%

  • 4.0/4.1 GHz

    Votes: 3 23.1%

  • 4.2/4.3 GHz

    Votes: 6 46.2%

  • 4.4/4.5 GHz

    Votes: 3 23.1%

  • 4.6/4.7 GHz

    Votes: 1 7.7%
  • Total voters
    13
  • Poll closed .
i6omg2K.jpg

Intel has released a pic of the Kaby Lake G NUC, beside a Apple Keyboard, a sign of things to come possibly?

Able to preorder the NUC here
 
a mainstream and affordable CFL 8 core. That would be futuristic that same and similar scaling.
 
oleNBR said:


worth the $20? about 3-4C under heavy loads
Click to expand...


No.. not really it's still gonna run hot. You need the ncore on die idea think about the only thing that could stop a lot of it. a chamber piece mold in shape of the die and sit on it perfectly.
 
Their 10nm is just a disaster, if they delay it again I think they will just skip it and work on 7nm. Goes to show just how much harder it's getting to shrink chips down much further.
 
No way will they 'skip it's. If anything, we will see 10nm, 10+nm, 10++nm, 10+++nm, etc.
Pretty sure we are getting 14+++ nm in the form of Whiskey Lake already. If that is the case, you can guarantee a 10++++nm as we will only stay on the nodes longer.
 
Dayman said:
Or maybe they'll co-exist instead of one being delayed! Intel said they'll be releasing more 8th Gen Chips in H2'18, not 9th Gen Chips, so this is still part of 8th Gen.

Guess we gotta wait and see :p
Click to expand...

yea man really hope so. 8 cores CFL 14nm++ aint so bad, though i am just hoping it is socket compatible and work with z370 cause dont want to purchase another laptop mobo.

Gideon said:
Their 10nm is just a disaster, if they delay it again I think they will just skip it and work on 7nm. Goes to show just how much harder it's getting to shrink chips down much further.
Click to expand...

even if it's a disaster, its already delayed for close to 2yrs now so if anything it should be out soon and since its almost out its worth waiting for because we'd get tiny IPC uplift, power efficiency over 14nm++ and more cores. and because its intel we have higher chance hitting 5ghz+. i know thats brag right and 5ghz isnt much different from AMD's 4.4ghz, that 600mhz i want it.
 
oleNBR said:
yea man really hope so. 8 cores CFL 14nm++ aint so bad, though i am just hoping it is socket compatible and work with z370 cause dont want to purchase another laptop mobo.



even if it's a disaster, its already delayed for close to 2yrs now so if anything it should be out soon and since its almost out its worth waiting for because we'd get tiny IPC uplift, power efficiency over 14nm++ and more cores. and because its intel we have higher chance hitting 5ghz+. i know thats brag right and 5ghz isnt much different from AMD's 4.4ghz, that 600mhz i want it.
Click to expand...

I think they are having problems with yields on anything large enough that you would want. They are using lots of masks to make 10nm and the chance of error gets pretty high, I honestly think you will have to wait for EUVL tech to be deployed before we see something for the desktop. By the time they deploy this, they just might move on to 7nm at that point. 14nm is working good enough for them for now so they may just skip 10nm if they are still having trouble with it.
 
TaintedSquirrel said:
Found on Reddit, pretty interesting. Someone a week or two ago was reporting similar temp results and I shrugged it off.

Click to expand...

yes I reported on here my temps are nowhere near as people have been reporting, but not many people have a 8600k, the 8700k is by far more popular so I put my lwoer temps down to the lack of HTT.
 
when intel says Q2 usually thats always end of the quarter right? 2H would be end of the yr even though its anything between June to December.

8 core CFL around end of June, there any sort of event to show this? i have never seen a CPU release around June time especially not mainstream CPU or this is another one of those "August CFL 6 core release but real date becomes December" type.
 
Skylake-x was released in June iirc but yeah those are not realy mainstream, though the original Skylake's were also 5th of august so there is always a chance.
 
oleNBR said:
when intel says Q2 usually thats always end of the quarter right? 2H would be end of the yr even though its anything between June to December.

8 core CFL around end of June, there any sort of event to show this? i have never seen a CPU release around June time especially not mainstream CPU or this is another one of those "August CFL 6 core release but real date becomes December" type.
Click to expand...
CY or FY is the important question, but I don't know what 360-day period Intel uses for their FY.
 
chrcoluk said:
yes I reported on here my temps are nowhere near as people have been reporting, but not many people have a 8600k, the 8700k is by far more popular so I put my lwoer temps down to the lack of HTT.
Click to expand...

I've seen 370 motherboards go a bit nuts on the voltages, presumably in hopes of always having a "it just worked" response for overclocking. While it works, it's hotter than hell.

While I don't ever expect a one-button setting on a mobo to compare to hand-tuned settings for a specific chip, this generation seems to be pretty significant in the typical vs worst case requirements. I dropped 20c on mine by setting voltages to a more reasonable level, and it's still not to the point where I'd consider it a golden chip really. It's just not the worst. In fact, the UEFI SVID response setting even calls this out - setting it to "typical" or "best case" makes a HUGE difference versus default of worst-case.

SMT does increase temps, but really only because it is allowing a higher utilization of the cores. That was actually the primary design goal - make sure execution units are not idle. But if you are really saturating all the cores, the difference becomes fairly irrelevant.
 
Nightfire said:
Over twice my OVERCLOCKED 2400g in gaming performance with similar cpu performance. Should be close to a ps4 pro with solid 1080p gaming at high. Damn impressive!
Click to expand...

It is impressive, but at what cost...

You can outfit a 2400g with motherboard/RAM/SSD for maybe 1/4 of the price. If Intel came in with these around $599, more people would consider it, but $999 for the barebone drives it into a niche category where people are paying specifically for the size. I mean an Asus ROG G20 with a 6700/GTX1070 can be had for $900 on ebay right now which would run circles around it with a larger form factor.

If there was something like an Alienware Alpha with this in it and priced around $599, it would sell much, much better.
 
Yeah, it's a demonstration of compute density and will be nice for the mobile platforms that eventually ship with it, but won't ever be really competitive on price/performance.
 
Big Release Day for Intel, eh?

CFL H, CFL U, More CFL S, The New Cannonlake 14nm Chipsets(B360, H370, H310, Q370 etc)
 
^^^ Funny, this feels like the time when Intel originally intended to launch CFL instead of the Z270 rebrand and no budget boards for 6 months...



But no, the original release wasn't a response to AMD according to some in this thread...a completely planned exercise to make their brand new 6 month old Z270 chipset obsolete, rebrand it as Z370, and then have no budget offerings for another 6 months.
 
Last edited:

excellent. now just hoping we'll be able to drop 8c into z370 with the same socket..
 
These mobile chips look really impressive. Has anyone seen a review mention their temps?

The 7700HQ was a pain to cool because of its size. If you had the GPU loaded and the heatsinks heat soaked, I keeping it under 90c could be hard (even while the gpu would happily run at 70-80 despite using 100w vs 40w)

Now it's at 10% higher frequencies which alone seems impossible.
 
jenosmaverick said:
Uhm are there any perf difference between a stock 8600k and 8600?
Click to expand...
It's locked and lower power compared to the -K part. At stock Turbo speeds there should be very little difference between the two. Anyone know if the -K parts still lack support for VT-d?
 
You must log in or register to reply here.
Back
Top