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 .
juanrga said:
The funny part is that this 10nm-is-delayed FUD is coming from the same people that promised us Zen2 on 7nm next year, despite AMD official roadmaps said otherwise.
Click to expand...

Maybe you are confused, 10nm predictions I have seen were Nvidia actually. They went from 10nm sampling last year, to not using it for Volta. Tricky node for them.

Tape out this year, so Q4 '18-Q1-'19 was not unrealistic expectation. Now with new information '12nm' Zen+ coming out Feb 2018 we know not to expect Zen2 in 2018.
Early-Mid 2019 is likely instead. Not a huge difference is it?
Wouldn't you rather have AMD improve Ryzen with a 'tock' prior to Zen2? Oh wait, no you wouldn't because anything AMD does is bad for your efforts here.
That said, 10, 12, 14, 16nm etc are mostly marketing terms these days, even on phones and really have much less relevance to the actual product at hand. Back in the days of 120nm, 90nm, 65nm etc yeah the process nodes made a big difference. These days the lines are blurred.

P.s. where is your icelake prediction blog? 4.7%!!!!11!
 
SomeGuy133 said:
they is no new delay...it was click bait sensalization. 10nm ULV laptop parts are on track as they "always" have been. 10nm desktops have been canned and replaced with coffee lake and possibly a future 8 core coffe lake instead of cannon lake with icelake to follow. on a process and arch step. Assuming intels charts are to be taken literal 10nm+ will be slightly weaker than 14nm ++ due to it not being as mature.




can both of you please stop this ridiculous sensationalization of everything...... it provides nothing useful to this thread and its just childish

also gideon stop making things up with 0 factual basis. it is getting tiresome to read and i rather not want to ignore you like ignoring abrasion

abrasion your making stuff up too and tons of PR and historical facts and scinfic data says what you said was beyond retard.

A lot of the info is released and your just spewing pointless ignorant stuff to new members and spreading patently false lies.
Click to expand...


https://www.electronicsweekly.com/n...cannonlake-10nm-processor-third-time-2017-09/
https://www.fool.com/investing/2017/07/26/the-price-of-intel-corporations-10-nanometer-failu.aspx
http://marketrealist.com/2017/03/could-intels-10nm-node-face-another-delay/
https://www.digitaltrends.com/computing/intel-10nm-cannonlake-schedule-rumors/

Yep looks like it's been all smooth for Intel.. Oh wait no it's been a disaster, unless you think all these places are making it up and I barely scratched at it. If you cant handle the truth go find a safe place.
 
Gideon said:
https://www.electronicsweekly.com/n...cannonlake-10nm-processor-third-time-2017-09/
https://www.fool.com/investing/2017/07/26/the-price-of-intel-corporations-10-nanometer-failu.aspx
http://marketrealist.com/2017/03/could-intels-10nm-node-face-another-delay/
https://www.digitaltrends.com/computing/intel-10nm-cannonlake-schedule-rumors/

Yep looks like it's been all smooth for Intel.. Oh wait no it's been a disaster, unless you think all these places are making it up and I barely scratched at it. If you cant handle the truth go find a safe place.
Click to expand...



Your last link literally says Intel has denied the delay...
 
The most recent delay has been denied too, even if it's only mobile in 2017, that's still 2017 for 10nm
 
Dayman said:
The most recent delay has been denied too, even if it's only mobile in 2017, that's still 2017 for 10nm
Click to expand...

Well we know it wont be desktop cause Intel finally admitted it was dead. They have a few more months to try to get that mobile chip out.
 
I think Coffee Lake is going to be my next build. Will finally be upgrading my 3930k, which I have to say has been a loyal and rock solid workhorse the past ... damn.. over 5 years. By and far, the most use I have gotten from a build. I will be going from 6 core to 6 core. I contemplated Ryzen, but single core performance on CL looks to be superior, the price is certainly reasonable and the platform is very stable. Contemplated x299 but overkill for what I need in a workstation and the price is just in orbit.
 
Gideon said:
Well we know it wont be desktop cause Intel finally admitted it was dead. They have a few more months to try to get that mobile chip out.
Click to expand...

Why would desktop get 10nm first? Also Cannon Lake have been for mobile only for ages. And being a 2+2 design.
 
Intel's own projections show that initial 10nm is only going to produce lower power chips and has for a long time now.
intel-xeon-process-technology-enhancements.jpg


It wouldn't make any sense for them to release desktop cpus on 10nm until its at 10nm+ and near parity with 14++ which is what Coffee lake is bringing finally.

Pretty sure if you look at intel's release schedule in recent times low power / mobile is first to new process nodes followed by consumer once they have brought the process up to speed on clocks and later on HEDT/Server once its fully matured. Makes prefect sense IMHO.

Besides at this stage with everyone's processes being slightly different in various ways I think its pretty pointless to put any weight at all into new process nodes. nobody knows how its really going to work in the real world till its a reality and in the wild and so far Intel has been managing to do some really great things with 14nm now that it has matured extensively, meanwhile you take a look at GloFlo/samsung 14nm its pretty far behind even skylake (14nm) on clocks now glflo/amd has decided to rename their 14+ to 12nm to get people excited but until we see Zen cpus clocking 4.3-4.5+ ghz on this process improvement who knows how good or bad it is only time will tell.
 
Gideon said:
Yep in February 2016 and they said it would release in 2017, yet here we are with the third 14nm launch from Intel.
Click to expand...

yea this goes to show intel can literally change anything if they wanted. there might really be a delay though. if they felt like it they could have come up with a statement to delay it to end of 2019 or 2020 if theres no competition and refresh 10nm for yrs. thanks to AMD they finally gave something decent, almost 10 yrs on quad cores and suddently 6-8 cores mainstream in less than 2 years and its not because of AMD? even monkey will understand its because of AMD.


Frito11 said:
Intel's own projections show that initial 10nm is only going to produce lower power chips and has for a long time now.
intel-xeon-process-technology-enhancements.jpg


It wouldn't make any sense for them to release desktop cpus on 10nm until its at 10nm+ and near parity with 14++ which is what Coffee lake is bringing finally.

Pretty sure if you look at intel's release schedule in recent times low power / mobile is first to new process nodes followed by consumer once they have brought the process up to speed on clocks and later on HEDT/Server once its fully matured. Makes prefect sense IMHO.

Besides at this stage with everyone's processes being slightly different in various ways I think its pretty pointless to put any weight at all into new process nodes. nobody knows how its really going to work in the real world till its a reality and in the wild and so far Intel has been managing to do some really great things with 14nm now that it has matured extensively, meanwhile you take a look at GloFlo/samsung 14nm its pretty far behind even skylake (14nm) on clocks now glflo/amd has decided to rename their 14+ to 12nm to get people excited but until we see Zen cpus clocking 4.3-4.5+ ghz on this process improvement who knows how good or bad it is only time will tell.
Click to expand...

that 10nm++ looks so far away man.

10nm+ is slightly behind 14nm++ so unless theres some sort of decent IPC jump 3-4% at least and decent power efficiency theres literally no point of buying it. of course intel knows this so thats why they put out 2more cores than CFL to entice people, along with newer chipset.
 
oleNBR said:
that 10nm++ looks so far away man.

10nm+ is slightly behind 14nm++ so unless theres some sort of decent IPC jump 3-4% at least and decent power efficiency theres literally no point of buying it. of course intel knows this so thats why they put out 2more cores than CFL to entice people, along with newer chipset.
Click to expand...

Indeed 14++ is likely to be Intel's best process for some time, this is why there is a lot of talk about TSMC, GloFlo and Samsung possibly getting ahead of Intel in the next few years with their 7nm process but the other problem with that is Intel's 10nm is more like 7nm than theirs are and that's what i'm getting at, at this stage in time process node size is no longer anything much more than marketing jumble to get people and investors excited. we're reaching the very limits of what can be done to reduce size and it just gets harder and harder to make it work properly the smaller it gets so until theres actual chips in the wild doing great things everything is speculation.
 
oleNBR said:
yea this goes to show intel can literally change anything if they wanted. there might really be a delay though. if they felt like it they could have come up with a statement to delay it to end of 2019 or 2020 if theres no competition and refresh 10nm for yrs. thanks to AMD they finally gave something decent, almost 10 yrs on quad cores and suddently 6-8 cores mainstream in less than 2 years and its not because of AMD? even monkey will understand its because of AMD.




that 10nm++ looks so far away man.

10nm+ is slightly behind 14nm++ so unless theres some sort of decent IPC jump 3-4% at least and decent power efficiency theres literally no point of buying it. of course intel knows this so thats why they put out 2more cores than CFL to entice people, along with newer chipset.
Click to expand...


I think your likely right on all points.
 
Frito11 said:
Indeed 14++ is likely to be Intel's best process for some time, this is why there is a lot of talk about TSMC, GloFlo and Samsung possibly getting ahead of Intel in the next few years with their 7nm process but the other problem with that is Intel's 10nm is more like 7nm than theirs are and that's what i'm getting at, at this stage in time process node size is no longer anything much more than marketing jumble to get people and investors excited. we're reaching the very limits of what can be done to reduce size and it just gets harder and harder to make it work properly the smaller it gets so until theres actual chips in the wild doing great things everything is speculation.
Click to expand...

Intel tried to do 10nm with just using masks and refining the process, it just didnt work like they hoped, so they are stuck waiting on EUV tech as well. Really no one cares if AMD 7nm or Intel 10nm is a smaller process, we will only care how the chip performs.
 
  • Like
Reactions: noko
like this
oleNBR said:
yea this goes to show intel can literally change anything if they wanted. there might really be a delay though. if they felt like it they could have come up with a statement to delay it to end of 2019 or 2020 if theres no competition and refresh 10nm for yrs. thanks to AMD they finally gave something decent, almost 10 yrs on quad cores and suddently 6-8 cores mainstream in less than 2 years and its not because of AMD? even monkey will understand its because of AMD.
Click to expand...

The ancient Westemere mainstream 'APU' was 2-core and SandyBridge introduced the first 4-core in year 2011, doubling the number of cores. Originally, Skylake (14nm) was going to be the last 4-core, and CannonLake (10nm) would increase the number of cores up to 8. This was the original plan. But then Intel did start having problems with 14nm and 10nm nodes (as every else foundry) and the whole plan was changed. The classic tick-tock strategy was replaced by tick-tock-optimization-2nd-optimization and after Skylake, we got first 4-core Kabylake (14nm+) and now 6-core CoffeLake (14nm++).

Once the 10nm problems are solved, we will get 8-core 'APU' back in the roadmap. Last rumor is that this is coming next year with IceLake.

On the other hand, AMD has been locked on quad-core APUs since 2011:

"Llano" (2011) --> 4-core
"Trinity" (2012) --> 4-core
"Richland" (2013) --> 4-core
"Kabini" (2014) --> 4-core
"Kaveri" (2014) --> 4-core
"Carrizo" (2016) --> 4-core
"Bristol Ridge" (2017) --> 4-core
"Raven Ridge" (2018) --> 4-core
"Picasso" (2019) --> 4-core

Paraphrasing you, we can hope that AMD will "finally gave something decent", such as a 6-core or 8-core APU before 2022. And when this happen we will say: thanks Intel!
 
kohl said:
I think Coffee Lake is going to be my next build. Will finally be upgrading my 3930k, which I have to say has been a loyal and rock solid workhorse the past ... damn.. over 5 years. By and far, the most use I have gotten from a build. I will be going from 6 core to 6 core. I contemplated Ryzen, but single core performance on CL looks to be superior, the price is certainly reasonable and the platform is very stable. Contemplated x299 but overkill for what I need in a workstation and the price is just in orbit.
Click to expand...

Yes CoffeeLake,..will be far improved single core. However, in multithreaded work loads the Ryzen will dominate by a longshot. AMD has really perfected SMT as of todays environment.
 
Last edited:
Gideon said:
Intel tried to do 10nm with just using masks and refining the process, it just didnt work like they hoped, so they are stuck waiting on EUV tech as well. Really no one cares if AMD 7nm or Intel 10nm is a smaller process, we will only care how the chip performs.
Click to expand...

That's pretty much what i said..
 
tangoseal said:
Yes CoffeeLake,...what a retarded as fuck naming scheme Intel and AMD use btw, will be far improved single core. However, in multithreaded work loads the Ryzen will dominate by a longshot. AMD has really perfected SMT as of todays environment.
Click to expand...

8C RyZen giving only ~10% more performance than 6C CoffeLake even in favorable benches as Cinebench is not what I call domination.

Also I don't see any SMT perfection, only see AMD getting more gains when adding a second thread because the Zen core cannot extract so much ILP from code as Intel does.
 
juanrga said:
8C RyZen giving only ~10% more performance than 6C CoffeLake even in favorable benches as Cinebench is not what I call domination.

Also I don't see any SMT perfection, only see AMD getting more gains when adding a second thread because the Zen core cannot extract so much ILP from code as Intel does.
Click to expand...

Well we will have to see once official bench marks are out. But your probably right.
 
OK, what would be the logical solution for somebody who wants to use more than 64 Gigs of RAM in a workstation, which is the limit on the mainstream platforms?
Single core performance seems good on Coffeelake, but isn`t it better to get X299 board with a 7800X, then overclock it and install more memory?
 
christal said:
OK, what would be the logical solution for somebody who wants to use more than 64 Gigs of RAM in a workstation, which is the limit on the mainstream platforms?
Single core performance seems good on Coffeelake, but isn`t it better to get X299 board with a 7800X, then overclock it and install more memory?
Click to expand...

If your looking for more memory HEDT/Server is the way, consumer platforms have not and probably will not ever need more than 64 Gb ram. on my gaming rig i've been at 16gb sense sandy bridge at least, i still don't see that changing anytime soon.
 
juanrga said:
8C RyZen giving only ~10% more performance than 6C CoffeLake even in favorable benches as Cinebench is not what I call domination.

Also I don't see any SMT perfection, only see AMD getting more gains when adding a second thread because the Zen core cannot extract so much ILP from code as Intel does.
Click to expand...
Maybe he meant domination in the price to performance. :)
 
juanrga said:
The ancient Westemere mainstream 'APU' was 2-core and SandyBridge introduced the first 4-core in year 2011, doubling the number of cores. Originally, Skylake (14nm) was going to be the last 4-core, and CannonLake (10nm) would increase the number of cores up to 8. This was the original plan. But then Intel did start having problems with 14nm and 10nm nodes (as every else foundry) and the whole plan was changed. The classic tick-tock strategy was replaced by tick-tock-optimization-2nd-optimization and after Skylake, we got first 4-core Kabylake (14nm+) and now 6-core CoffeLake (14nm++).

Once the 10nm problems are solved, we will get 8-core 'APU' back in the roadmap. Last rumor is that this is coming next year with IceLake.

On the other hand, AMD has been locked on quad-core APUs since 2011:

"Llano" (2011) --> 4-core
"Trinity" (2012) --> 4-core
"Richland" (2013) --> 4-core
"Kabini" (2014) --> 4-core
"Kaveri" (2014) --> 4-core
"Carrizo" (2016) --> 4-core
"Bristol Ridge" (2017) --> 4-core
"Raven Ridge" (2018) --> 4-core
"Picasso" (2019) --> 4-core

Paraphrasing you, we can hope that AMD will "finally gave something decent", such as a 6-core or 8-core APU before 2022. And when this happen we will say: thanks Intel!
Click to expand...

lmao using APU in a laptop to justify intel 4 core desktop, theres a reason why laptop will overheat and they dont try to put more than 4 cores. to get 6 or 8 cores u'll need a fat giant cpu heatsink like the one in my p570wm with 1680v2 xeon. but no, just. simply. retarded. comparison.

intel could have kept laptop 4c due to heat and give desktop early access to 6-8 cores but NOOOOOO. i like intel cpu because its faster but im not blind like fanbois.

once again, thanks to AMD we finally got 6-8core mainstream desktop cpu, then when clevo finally put those in i can get 8 intel cores in a laptop with all the latest chipset/feature.
 
oleNBR said:
lmao using APU in a laptop to justify intel 4 core desktop, theres a reason why laptop will overheat and they dont try to put more than 4 cores. to get 6 or 8 cores u'll need a fat giant cpu heatsink like the one in my p570wm with 1680v2 xeon. but no, just. simply. retarded. comparison.

intel could have kept laptop 4c due to heat and give desktop early access to 6-8 cores but NOOOOOO. i like intel cpu because its faster but im not blind like fanbois.

once again, thanks to AMD we finally got 6-8core mainstream desktop cpu, then when clevo finally put those in i can get 8 intel cores in a laptop with all the latest chipset/feature.
Click to expand...


Intel had been planning to up core counts with coffee lake, at best AMD just forced them to release these sooner as they were originally slated to release Q1 next year along with Z390, now the rumored 8 core variant for next year if true i would say is probably a direct reaction to Ryzen long term wise.

Competition is indeed great and i've always been happy that Zen has been a success, heck i even was considering buying a R7 earlier this year but decided i didn't want to take an IPC hit to gain more cores which for the most part don't matter to me in my day to day use, 6/12 is plenty, 8/16 would be overkill.
 
Hagrid said:
Maybe he meant domination in the price to performance. :)
Click to expand...

I guess he did mean performance

https://hardforum.com/threads/intel...a-1151-6c-12t.1930226/page-56#post-1043240671

but even in your favorite metric, $500 RyZen chip giving ~10% more performance than a $360 CFL chip on AMD-favorable benches doesn't look as "domination" to me.

oleNBR said:
lmao using APU in a laptop to justify intel 4 core desktop, theres a reason why laptop will overheat and they dont try to put more than 4 cores. to get 6 or 8 cores u'll need a fat giant cpu heatsink like the one in my p570wm with 1680v2 xeon. but no, just. simply. retarded. comparison.

intel could have kept laptop 4c due to heat and give desktop early access to 6-8 cores but NOOOOOO. i like intel cpu because its faster but im not blind like fanbois.

once again, thanks to AMD we finally got 6-8core mainstream desktop cpu, then when clevo finally put those in i can get 8 intel cores in a laptop with all the latest chipset/feature.
Click to expand...

I don't know why you mention laptops. I only discussed desktop.

Intel is giving 6-core 'APU' now and 8-core are in the future plans. Thanks Intel!

On the other hand AMD continues giving 4-core APUs for desktop even in 2019. After a decade AMD is still giving quad-cores for mainstream. Without Intel, AMD would be still releasing single core CPUs. Thanks Intel!
 
Frito11 said:
Intel had been planning to up core counts with coffee lake, at best AMD just forced them to release these sooner as they were originally slated to release Q1 next year along with Z390, now the rumored 8 core variant for next year if true i would say is probably a direct reaction to Ryzen long term wise.

Competition is indeed great and i've always been happy that Zen has been a success, heck i even was considering buying a R7 earlier this year but decided i didn't want to take an IPC hit to gain more cores which for the most part don't matter to me in my day to day use, 6/12 is plenty, 8/16 would be overkill.
Click to expand...

once u have had 8c its hard to go back, well only if you really have use for it. ipc is important same with frequency thats only because majority of software including OS are still not optimized, once they are, more cores are better.


juanrga said:
I guess he did mean performance

https://hardforum.com/threads/intel...a-1151-6c-12t.1930226/page-56#post-1043240671

but even in your favorite metric, $500 RyZen chip giving ~10% more performance than a $360 CFL chip on AMD-favorable benches doesn't look as "domination" to me.



I don't know why you mention laptops. I only discussed desktop.

Intel is giving 6-core 'APU' now and 8-core are in the future plans. Thanks Intel!

On the other hand AMD continues giving 4-core APUs for desktop even in 2019. After a decade AMD is still giving quad-cores for mainstream. Without Intel, AMD would be still releasing single core CPUs. Thanks Intel!
Click to expand...

why not APU in laptop? its not like you're stupid or something. prior to ryzen came out u know as well as i do their CPU is power hungry. no point making 12 or 16 core cpu when competition just have you stomped.

but what we got here as soon as ryzen came out we see 6 and 8 cores laptop housing them. intel could have easily up their mainstream core counts as early as 2 yrs ago but didn't, which is a real shame. they purposely delay it and delay it some more until they have a real need to. broadwell was already a good genera nation with decent efficiency, could up it to 6c already.
 
Newegg reviews if each r7 cpu:
237 - 1700, 124 - 1700x, 98 - 1800x
On Amazon:
323- 1700, 121 - 1700x, 168 - 1800x

Does anyone still think most are buyinhg1800x over the rest of R7?

Price of 1800x: newgg - $450, amazon $420. NOT $500 as stated here.

Of course everyone on [H] knows that an R7 1700 can do the same for $300.

Maybe we can start quoting Xeon prices for Kabylake....
 
Your original argument was about CoffeLake 'APU' and RyZen CPUs for desktop. After I demonstrated the history behind Intel roadmaps is different, and that AMD APUs are locked on quad-cores up to 2019, now you want to change the discussion to laptops. Why I have a feeling that if I start discussing laptops you will change the argument again?
 
Nightfire said:
Newegg reviews if each r7 cpu:
237 - 1700, 124 - 1700x, 98 - 1800x
On Amazon:
323- 1700, 121 - 1700x, 168 - 1800x

Does anyone still think most are buyinhg1800x over the rest of R7?

Price of 1800x: newgg - $450, amazon $420. NOT $500 as stated here.

Of course everyone on [H] knows that an R7 1700 can do the same for $300.

Maybe we can start quoting Xeon prices for Kabylake....
Click to expand...

Everyone knows that cheaper products sell more. R7 1700 is worse die and overclocks worse. Prices mentioned are launch. 10% offers/discounts over launch price don't change the main point about how CoffeLake is a better product than RyZen series and AMD will be forced to apply huge discounts.
 
Max o/c for all r7 chips have been 3.8 to 4.1 ghz. There isnt any special sauce in the 1800x.

A $300 1700 will do just fine against a $360 8700k. A 1700x at the same price will do fine as well considering the geekbench and cinebench scores.
 
The first 4 reviews of a 1700 showed 4.0, 3.9, 3.9, and 4.0 ghz. The 1800x MIGHT have a 100 mhz advantage.
 
oleNBR said:
once u have had 8c its hard to go back, well only if you really have use for it. ipc is important same with frequency thats only because majority of software including OS are still not optimized, once they are, more cores are better.
Click to expand...

Yeah you know i've been building PC's sense pentium 75mhz when i a teenager, the myth that everything will take advantage of tons of cores died back in the dual and esp quad core days. it took AGES for anything that could be optimized to get optimized to use two and later on 4 cores. yes these days we are just now starting to see some games use 6-8 threads and beyond effectively but fact will long remain that things will be stuck in the past because software devs esp for games will be lazy and older stuff won't get re-worked to perform better. for gaming esp. clocks and IPC have always been and i'm sure always will be king. if your into redering/ transcoding / other computing type tasks or running your own servers at home sure you love the cores but for typical users and gamers it doesn't matter till enough game devs make it matter and that can take many years.
 
Were approaching the limits of physics when we get around 5nm... anything smaller might be a disaster. I am not sure what is next but I am not an engineer.
 
this makes me curious, looks like core architecture cant get optimized anymore so no more "core" ipc increase. when will we expect to see a new build soon from intel? AMD's got ryzen, hope intel can bring some nice things to the table, something like actual 10% ipc increase in a single generation due to architecture change would be nice.
 
oleNBR said:
this makes me curious, looks like core architecture cant get optimized anymore so no more "core" ipc increase. when will we expect to see a new build soon from intel? AMD's got ryzen, hope intel can bring some nice things to the table, something like actual 10% ipc increase in a single generation due to architecture change would be nice.
Click to expand...

What's your source on Intel not being able to increase IPC?
 
oleNBR said:
this makes me curious, looks like core architecture cant get optimized anymore so no more "core" ipc increase. when will we expect to see a new build soon from intel? AMD's got ryzen, hope intel can bring some nice things to the table, something like actual 10% ipc increase in a single generation due to architecture change would be nice.
Click to expand...


Skylake, Kaby and Coffee lake are all the same skylake core on refined nodes and more cores, wait till Icelake (new core) to make the judgement of no IPC increases....
 
You must log in or register to reply here.
Back
Top