I think if it's standard Coffee Lake architecture, the base clock will end up being too low.
They funny about this is that people wrote the same about Coffeelake 6C. Base clock doesn't really matter for i7-8700K. Same for the temp worries.
I think if it's standard Coffee Lake architecture, the base clock will end up being too low.
It probably will matter with the 2 extra cores and 4 extra threads.They funny about this is that people wrote the same about Coffeelake 6C. Base clock doesn't really matter for i7-8700K. Same for the temp worries.
2H18 is not far away and there really haven't been any solid 9700K or Z390 leaks yet, have there?
It probably will matter with the 2 extra cores and 4 extra threads.
Or the Intel way of keeping "95 W" (more like 125 W in practice), but also 4.3-4.8 GHz.
This is exactly the same what people told before CFL 6C came out. In the end what matters in almost all cases is the Turbo. Base clock is important for extreme AVX workloads which isn't really a topic beside stress tests like Prime95. Maybe they have to go down from 4.3 to 4.0 Ghz for the All Core Turbo on CFL 8C. They might also think about a TDP increase to 105W following AMDs route.
Actually the 8700K draws almost exactly 95W (in Blender) according to power draw measurements done by Gamers Nexus: https://www.gamersnexus.net/hwrevie...vs-ryzen-streaming-gaming-overclocking/page-3
Ahh fair enough. Then I would agree, would be difficult to maintain 95W unless lower clocks or voltages are usedTo clarify, in my original comment, I suggested to spec Coffee Lake 8-core as 4.3-4.8 GHz, 125 W (rated "95 W").
People still not realizing that the "mainstream" line of chips are designed for laptop and that the desktop gets whatever the CPU team can get away with overclocking the laptop chips to.
8 core coffee lake would be an absolutely worthless laptop chip at 35w-45w (Intel's preferred mainstream laptop TDP) compared to a 6 core coffee lake in the same TDP.
390 is the "cannonlake" chipset, which means it would support "icelake" chips as well.
If the rumor has any truth to it at all, it's early engineering samples for icelake cropping up on cannonlake chipset.
Intel doesn't make desktop chips.
They make "mainstream" laptop chips that go into most desktops.
They make "high core count server" chips that go into some desktops.
Ryzen (8c+) also isn't a "desktop" chip, it's a "high core count server" chip re-purposed for desktops.
Intel hasn't made a "desktop" chip since Pentium 4, so stop this stupid argumentation point.
Damn, I had no idea I've been running laptop CPUs on my desktop motherboard for the past decade and a half. Argument over then, Intel doesn't make desktop CPUs...
You are just another board poster as far as I can tell. Why should I rate your claims above any one else's?Keep believing that Intel would make a mask and an entire "mainstream" line-up purely for desktop that would be terrible for laptop with your only "source" being random AMD viral marketers then.
The total lack of any technical/financial knowledge left on these forums other than the few random AMD engineers assigned to PR posting is showing again.
You are just another board poster as far as I can tell. Why should I rate your claims above any one else's?
"random AMD viral marketers"
"AMD engineers assigned to PR posting"
I've been an Intel fan for a very long time, and that sort of line always undermines credibility...
You might want to leave out the fanboy type remarks in the future. No one will take you seriously, except possibly members of the choir, and that's of no use to you.
8 core coffee lake would be an absolutely worthless laptop chip at 35w-45w (Intel's preferred mainstream laptop TDP) compared to a 6 core coffee lake in the same TDP.
Intel has in the past done 55/57W "mobile" parts, that could use the 8 core as well. The 9th Gen 35-45 W H could remain 6 core Coffee Lake.
The problem of course is that 10 nm yield continues to be horrific, AND top clocks of Icelake are going to be lower enough that a desktop part will be slightly slower in games. Combine the two, and the need to release something new every year as to why they would do the 8C CFL.
All that matter is what Intel thinks.6 core mobile coffee lake isn't even out yet.
6 core mobile coffee lake is the chip you are thinking about if you are talking about a "laptop stop-gap for 2018"
It's hilarious that this forum thinks that Intel is "in dire need of a 8 core coffee-lake to counter the unstoppable Ryzen".
Ryzen caught AMD up to Sandy Bridge, and they fired their CPU designer, so that means it's Bulldozer vs Sandy Bridge all over again, as Skylake is roughly as far ahead of Ryzen as Sandy Bridge was ahead of Bulldozer.
6 core mobile coffee lake isn't even out yet.
6 core mobile coffee lake is the chip you are thinking about if you are talking about a "laptop stop-gap for 2018"
That last bit is bananas...
Yep, nowhere near the gap.
Cinebench singlethreaded, Sandy Bridge had a 50% advantage over 1st gen Bulldozer CPUs, 60% IPC but Bulldozer was clocked higher. By comparison An 8700K has a 26% advantage over a 1700X, and probably 10% of that is clockspeed. Software is better threaded now so the extra cores in Ryzen have more value. I still wouldn't buy Ryzen, it doesn't fit my use-case, but it's not Bulldozer over again.
Cinebench was never a good "benchmark" for anything, which is why it's brought up all the time for bad comparisons (It's an obsolete GPU benchmark from forever ago).
In actual applications that people use CPUs (not GPUs) for, it's the same gap.
Cinebench was never a good "benchmark" for anything, which is why it's brought up all the time for bad comparisons (It's an obsolete GPU benchmark from forever ago).
In actual applications that people use CPUs (not GPUs) for, it's the same gap.
I would be interested to see which applications are running 50-60% faster on CFL than Ryzen, with significantly higher power usage to boot. Not that I actively disbelieve you, I just haven't yet seen this.