Coffeelake thread, benchmarks, reviews, input, everything.

[jpiniero]

A launch at Gamecom would make sense. Given that Intel is apparently not changing much other than clocks besides the 8 core i9 there isn't much reason to hold it back... unless the next desktop chip is more likely to be released in early 2020 and so they didn't want the time gap to be that large. Remember that Coffee Lake's launch was supposed to be earlier this year but they pulled it in because of Ryzen.

 

[mikk]

A launch at Gamecom would make sense. Given that Intel is apparently not changing much other than clocks besides the 8 core i9 there isn't much reason to hold it back...

Who is saying they are holding back? If it isn't ready they can't launch unless you want a paper launch.

 

[jpiniero]

Who is saying they are holding back? If it isn't ready they can't launch unless you want a paper launch.

That's fair. Capacity issues are a problem I imagine which I think may have encouraged Intel to limit the number of SKUs using the 8 core die (for now anyway).

 

[LTC8K6]

Who is saying they are holding back? If it isn't ready they can't launch unless you want a paper launch.

At this point, I'll take a paper launch.

 

[epsilon84]

Z370 support for Intel's Core i9 Coffee Lake 8-core now confirmed
https://www.pcgamesn.com/intel-core-i9-z370-support

Kind of surprised no one has picked up on this. Good news for 8th gen CFL owners running Z370 mobos, though I have no plans to move on my 8700K anytime soon. Guess the Z370 platform is not so 'dead end' after all... will be interesting to see how Z390 differentiates itself though.

 

[jpiniero]

Guess the Z370 platform is not so 'dead end' after all...

Does mean that Icelake will be on a new socket though, if they manage to fix 10 nm and end up releasing desktop parts.

 

[PingSpike]

Z370 support for Intel's Core i9 Coffee Lake 8-core now confirmed
https://www.pcgamesn.com/intel-core-i9-z370-support

Kind of surprised no one has picked up on this. Good news for 8th gen CFL owners running Z370 mobos, though I have no plans to move on my 8700K anytime soon. Guess the Z370 platform is not so 'dead end' after all... will be interesting to see how Z390 differentiates itself though.

I don't really feel that headline matches the article, but its a good roundup of the evidence to date. The evidence they cite is the cpu id codes from the MSI bios update. This *is* the strongest evidence we have of it happening and its bolstered by ASUS also having the cpu id code in a bios update, but its not definitive or official from Intel. At least, not yet anyway.

All I've seen about z390 is it adds on chipset wifi and I think BT and extra USB 3.1 ports? Not willing to pay extra personally since I don't use wifi on my desktops.

 

[epsilon84]

I don't really feel that headline matches the article, but its a good roundup of the evidence to date. The evidence they cite is the cpu id codes from the MSI bios update. This *is* the strongest evidence we have of it happening and its bolstered by ASUS also having the cpu id code in a bios update, but its not definitive or official from Intel. At least, not yet anyway.

All I've seen about z390 is it adds on chipset wifi and I think BT and extra USB 3.1 ports? Not willing to pay extra personally since I don't use wifi on my desktops.

Yeah fair points, it's not an official announcement but if that ID code is for an 8 core CPU then what else could it be?

 

[mikk]

Z370 support for Intel's Core i9 Coffee Lake 8-core now confirmed
https://www.pcgamesn.com/intel-core-i9-z370-support

Kind of surprised no one has picked up on this. Good news for 8th gen CFL owners running Z370 mobos, though I have no plans to move on my 8700K anytime soon. Guess the Z370 platform is not so 'dead end' after all... will be interesting to see how Z390 differentiates itself though.

There is nothing new in this article. This is old news.

 

[Eug]

Judging by all the reports out there, I don't see how Intel's Core i9-8950HK TDP of 45 Watts makes any sense.

It's far, far too easy induce hard-core throttling in the chip with sustained loads with even just regular applications (not synthetic loads). The same machines can handle continuous 45 Watt loads just fine with zero throttling.

 

[StinkyPinky]

My understanding is that the 8 core will still be called 8th gen, so it would be ridiculous for it to not work in the Z370. It would be a great way to burn out what little goodwill Intel have left.

 

[epsilon84]

There is nothing new in this article. This is old news.

My bad, wasn't aware that upcoming 8 core CFL chips working on Z370 was old news. Was it already posted here?

 

[coercitiv]

My bad, wasn't aware that upcoming 8 core CFL chips working on Z370 was old news. Was it already posted here?

There were posts showing software support at BIOS level for up to 8c CPUs on Z370.

 

[mikk]

My bad, wasn't aware that upcoming 8 core CFL chips working on Z370 was old news. Was it already posted here?

Yes it was.

http://www.portvapes.co.uk/?id=Latest-exam-1Z0-876-Dumps&exid=thread...nput-everything.2519949/page-61#post-39498538

 

[jpiniero]

http://www.coolaler.com/threads/intel-core-i9-9900k_i7-9700k_i5-9600k.351755/

New rumor... 9700k might be 8C8T? Also ACT of 4.7 for the 9900k.

 

[Dayman1225]

Also ACT of 4.7 for the 9900k.

Unless the soldered rumor is true, there is no way in hell thats hitting 4.7 on all 8c at stock settings (imagine the auto voltage!)

EDIT: it appears to be fake, Official Intel documents listed 9600Ks turbo as 4.5GHz, not 4.6GHz

 

[jpiniero]

Unless the soldered rumor is true, there is no way in hell thats hitting 4.7 on all 8c at stock settings (imagine the auto voltage!)

The 8 core could be a new stepping I guess, or tighter binned on voltage.Plus ACT is not guaranteed, only that it is possible.

EDIT: it appears to be fake, Official Intel documents listed 9600Ks turbo as 4.5GHz, not 4.6GHz

Of course until Intel formally announces the products they are free to change the specs as need be. But yeah.

 

[mikk]

Intel would call a 8/8 CPU differently. Because 8700k-->9700k = 6/12

So pretty sure this is a fake.

 

[jpiniero]

Intel would call a 8/8 CPU differently. Because 8700k-->9700k = 6/12

If there is a tangible benefit that the 8 core die has that the others don't, it would make sense to do it this way.

 

[happy medium]

Intel 9th Generation Flagship, Core i9-9900K 8 Core, 16 Thread CPU Specifications Leaked –

4.7 all core turbo.

https://wccftech.com/intel-9th-gen-core-i9-9900k-i7-9700k-i5-9600k-cpu-specs-leak/

 

[Shivansps]

This is stupid, only 1 sku with HT?

 

[epsilon84]

8/8 doesnt make sense to me, I would think in most MT workloads a 6/12 8700K would beat an 8/8 9700K, though if the reported 4.6GHz ACT is true then that would go some way to closing the gap in MT throughput

 

[PeterScott]

8/8 doesnt make sense to me, I would think in most MT workloads a 6/12 8700K would beat an 8/8 9700K, though if the reported 4.6GHz ACT is true then that would go some way to closing the gap in MT throughput

It would be close, but clock for clock I would give the edge to 8/8. But it will be so close that Intel may want to avoid that comparison.

At best Intel HT delivers about 25% performance boost on fully parallel workloads. Something like Cinebench is around the best case for HT:

About 25% boost in both cases.

So to equate them setting the value of HT core to 1.25 and a non HT core to 1.0:

6c *1.25 (HT) = 7.5
8c * 1.0(No HT) = 8.0

8/8 has a slight advantage over 6/12.

 

[StinkyPinky]

Intel 9th Generation Flagship, Core i9-9900K 8 Core, 16 Thread CPU Specifications Leaked –

4.7 all core turbo.

https://wccftech.com/intel-9th-gen-core-i9-9900k-i7-9700k-i5-9600k-cpu-specs-leak/

Pretty amazing if true. 4.7 out the box.

So the next question, what is better? 8/8 or 6/12?

 

[epsilon84]

It would be close, but clock for clock I would give the edge to 8/8. But it will be so close that Intel may want to avoid that comparison.

At best Intel HT delivers about 25% performance boost on fully parallel workloads. Something like Cinebench is around the best case for HT:

About 25% boost in both cases.

So to equate them setting the value of HT core to 1.25 and a non HT core to 1.0:

6c *1.25 (HT) = 7.5
8c * 1.0(No HT) = 8.0

8/8 has a slight advantage over 6/12.

Good points, but I don't think cinebench represents the best case scenario for HT/SMT scaling.

For example, in Blender a 8700K is 47% faster than a 8600K. Even a 7700K beats the 8600K and that is with a 50% core deficit. Clock for clock the 8600K edges a 7700K but only barely. A 8700K only has 33% less cores than a 9700K.

https://www.gamersnexus.net/hwreviews/3157-intel-i5-8600k-review-overclocking-vs-8700k-8400/page-3

In Blender at least, based on this a 8700K would beat a 9700K if it is indeed an 8/8 chip. Other apps and benchmarks will be a mixed bag probably, like your CB example