Intel 10th Generation Core Case-badges Revealed

[btarunr]

Intel laid rest to speculation that post 9th generation, it could replace its Core brand with something else. The 10th generation Core processors, built around the 10 nm "Ice Lake" microachitecture, will feature the first noteworthy IPC increments since "Skylake" thanks to their new "Sunny Cove" CPU cores. These will also feature DLBoost, a fixed-function matrix-multiplication hardware that speeds up deep-neural net building and training by 5x, as well as certain AVX-512 instructions. The cores will be optimized to cope with 2.4 Gbps 802.11ax Wi-Fi and faster Ethernet standards. The first of these chips will target mobile computing platforms, and will be quad-core parts like the dies pictured below. To save notebook PCB real-estate, Intel will put the processor and PCH dies into a multi-chip module. It will be quite a wait for the desktop implementation, but at least you know what their case-badges look like.



View at TechPowerUp Main Site

 

[krykry]

I'm appalled how they don't realize that such a move will be a catastrophe. Putting out low core count, mobile CPU as the representative of their 10th generation CPUs will work just the opposite they think it will.

If they want investors and consumers to stay with them, they need nothing less but a new architecture that would soundly beat Ryzen 3xxx series.

 

[Vayra86]

The fact this is a headline says it all, really.

 

[ncrs]

I'm appalled how they don't realize that such a move will be a catastrophe. Putting out low core count, mobile CPU as the representative of their 10th generation CPUs will work just the opposite they think it will.

If they want investors and consumers to stay with them, they need nothing less but a new architecture that would soundly beat Ryzen 3xxx series.

This is not the first time they put out a broken 10nm product just as a smoke screen for investors. The first one was i3-8121U - a dual-core processor that was so broken it had to have the iGPU disabled. According to rumors the yields for this model were below 1%. It cost Intel millions to make at most thousands of "full functioning" ones.

 

[Landcross]

I feel like Intel is doing as much as they can to push articles about AMD away. I feel like there have been much much more Intel articles last days than AMD articles, while Intel has barely any product to launch until the end of this year. I haven't fact checked that though, just a feeling so take it with a grain of salt

 

[champsilva]

I'm appalled how they don't realize that such a move will be a catastrophe. Putting out low core count, mobile CPU as the representative of their 10th generation CPUs will work just the opposite they think it will.

If they want investors and consumers to stay with them, they need nothing less but a new architecture that would soundly beat Ryzen 3xxx series.

I don't know if you know, but CPUs for laptops sell more than desktop and AMD doesnot have anything new in this scenario?

 

[ppn]

The notebook market allows them to rebadge I3 as an I7 for more profit. How big is this thing, 8 core part will take only 20% more chip real estate to produce and 2 click copy paste operation. So now we wait. 20% IPC gain is quite something.

 

[BorgOvermind]

2000 - 180 nm
2002 - 130 nm
2004 - 90 nm
2006 - 65 nm
2008 - 45 nm
2010 - 32 nm
2012 - 22 nm
2014 - 14 nm -------|
2016 - 14 nm -------| ---> FAIL
2018 - 14 nm -------|

 

[chaosmassive]

WAKE UP, Intel
come on get back up and fight that AMD !

you know, I just want you guys fight and beat each other to death so I get either your CPU or AMD for cheap

 

[EarthDog]

2000 - 180 nm
2002 - 130 nm
2004 - 90 nm
2006 - 65 nm
2008 - 45 nm
2010 - 32 nm
2012 - 22 nm
2014 - 14 nm -------|
2016 - 14 nm -------| ---> FAIL
2018 - 14 nm -------|

Yeah, huuuuuuuuuuuge fail... until this moment, dominance in the market (performance wise) is now a fail? Nice.

 

[I No]

2000 - 180 nm
2002 - 130 nm
2004 - 90 nm
2006 - 65 nm
2008 - 45 nm
2010 - 32 nm
2012 - 22 nm
2014 - 14 nm -------|
2016 - 14 nm -------| ---> FAIL
2018 - 14 nm -------|

Very very very weak trolling mate.

 

[Vayra86]

2000 - 180 nm
2002 - 130 nm
2004 - 90 nm
2006 - 65 nm
2008 - 45 nm
2010 - 32 nm
2012 - 22 nm
2014 - 14 nm -------|
2016 - 14 nm -------| ---> FAIL
2018 - 14 nm -------|

Yeah its very strange that when complexity has increased over ten-fold, the R&D phase takes a bit longer. Very strange indeed. I smell a conspiracy. Aliens! THE BORG!

 

[Basard]

Holy shit, them are some nice stickers!

 

[dirtyferret]

OMG! OMG! Intel's theoretical 18% IPC increase totally pwnd's AMD's theoretical 15% IPC increase....

 

[Nkd]

OMG! OMG! Intel's theoretical 18% IPC increase totally pwnd's AMD's theoretical 15% IPC increase....

AMD was spot on with their IPC with zen and they will likely be spot on with zen 2.

 

[kings]

No surprise seeing Intel focus 10nm chips on laptops first!

This market is much bigger than desktop PCs and it is where AMD has less representation.

 

[TheGuruStud]

1.8% confirmed

 

[Imsochobo]

No surprise seeing Intel focus 10nm chips on laptops first!

This market is much bigger than desktop PCs and it is where AMD has less representation.

Also 10NM have less performance than their 14NM.
They have said this countless times, they cannot replace 14NM with 10nm in desktops because who would buy a 8700K with 20% less performance ?

 

[HwGeek]

Indeed, they went down from 4.8Ghz boos t[ i7-8665U ] to 4.1Ghz boost at same power envelop [25W/28W] so no 9900KS replacement for now.

 

[TheGuruStud]

Indeed, they went down from 4.8Ghz boos t[ i7-8665U ] to 4.1Ghz boost at same power envelop [25W/28W] so no 9900KS replacement for now.

Wait till you find out what the base clocks are (which is what these chips actually run at when tdp constrained). I predict they're stupid low. 10nm is just unusable.

 

[Tomgang]

10 NM. About dam time intel got something new out. But they are still behind as amd is on 7 NM. but i guess its better than nothing.

But it still nothing i am impressed by. Still quad-core yawn. Wake up intel, its 2019 and not 2012. Quad-core cant really impresse any one these days.

 

[HwGeek]

Wait till you find out what the base clocks are (which is what these chips actually run at when tdp constrained). I predict they're stupid low. 10nm is just unusable.

Indeed, since good shrink should gave them more core and more clock at same TDP, like 3900X they got +300Mhz boost clock and 50% more cores at same TDP of 2700X.

 

[Manu_PT]

Intel is not worried at all. It would be as simple as dropping prices. Imagine dropping 9900k to 400€ and 9700k to 300€. They dont do it because they dont need to. They have the best performance.

 

[TheGuruStud]

Intel is not worried at all. It would be as simple as dropping prices. Imagine dropping 9900k to 400€ and 9700k to 300€. They dont do it because they dont need to. They have the best performance.

Incorrect. They don't lower prices, b/c there's still enough idiots paying the Intel tax.

 

[Metroid]

Incorrect. They don't lower prices, b/c there's still enough idiots paying the Intel tax.

Well business wise they dont do it just yet because ryzen 3000 has not been released, things will change in about 8 weeks or so. AMD have done the same thing in 2006, I myself see the 9900k a fail, the 9700k a step in the right direction, hyperthreading must die but I dont see it happening now. As soon as we get around 16/32 cores, hyperthreading will be less and less used. Hyperthreading adds 50% more heat for 25% more performance + security issues. Is that good? only you can decide that, I myself dont think so.

Plus, Intel and AMD have been indirectly charging you $100 for hyperthreading/smt, know why? because it comprises 25% more multithreading performance, for example, the 9700k and 9900k.



The difference here is 342 points, that is 20 - 25% or so performance for hyperthreading.

Interesting enough reviews even here at tpu says that the 9900k consumes only 10% more power, in my and many other tests around the world differs from these reviews in that sense.

Do you really believe the 9900k consumes only 18 watts more than the 9700k, 100% multithreading? test it on intelburntest and see yourself your house burn together hehe

I made tons of tests with nehalen when hyperthreading came to exist, i7 920 used 350 watts with hyperthreading, without hyperthreading, around 180 watts and the same 25% multithreading performance that was at that time and the interesting thing was that, hyperthreading off my i7 920 4.2ghz used 1.18v stable 24/7, hyperthreading on used 1.40v stable. So I dont believe these numbers reviews throw on people regarding power consumption while hyperthreading is enabled.