We're now squishing even more on a 14nm die! (Just another day at Intel) Comet Lake!

[Arika]

The only thing worse than Intel's node size issues is their product code names

[CarlBar]

4 hours ago, Misanthrope said:

Do other fabs have a non-compete signed with AMD? Because it's looking more and more like they'll have to just bite the bullet and use someone else's fabs at one point: they really don't want to let AMD get the unquestionable upper hand and majority market share.

 

Intel's 14nm process is basically on par with most other's 10nm process. Intel processes generally have higher logic density at a given node size.

 

3 hours ago, The Benjamins said:

This makes me think Zen2 mainstream has to be more then 8 cores. Look at how Intel is panicing and rushing to get more cores out.

 

Yeah, i pointed this out a few times now but in theory given everything AMD has told us if they were willing to settle for 9900k like thermals they could, (based on what AMD has revealed to us and whats leaked), build a 16 core chip with a 4.4Ghz base, 5.2Ghz Boost. That would be pushing the silicon and thermals to their absolute limits, but they could do it. 

 

1 hour ago, porina said:

I have a 14 core Intel CPU at 22nm, it doesn't melt, although its clocks reflect that.

 

The mainstream core war is going into silly territory now. 8 cores is already pushing dual channel ram. If either side go beyond that... it is getting really silly. A fat L4 might help but it is more a quick fix than a real solution, which is quad channel ram. At this point, just bring HEDT pricing down.

 

I wouldn't be surprised to see Ryzen get a quad channel memory controller that can function as dual channel for backwards compatibility.

 

I can't help but wonder if this might not be a pair of golden sample 8700K's on the same die with the worst core disabled, HT disabled, and a new turbo boost that can max out upto 4 cores. Based on  pre-binning some retailers did of the 8086 intel might be able to squeeze 5.3Ghz boost clocks out of the silicon. It wouldn't run any hotter than the 9900K, and the performance boost would be marginal at best, but it would give Intel a nice marketing trick.

[porina]

18 minutes ago, CarlBar said:

I wouldn't be surprised to see Ryzen get a quad channel memory controller that can function as dual channel for backwards compatibility.

It's called threadripper. They can't radically change AM4.

[Tedster]

1 hour ago, Dabombinable said:

Higher density=higher chance of having to disable faulty cores if I remember correctly. Also it means a higher cost.

Will we see Intel selling odd numbers of cores (other than one) in machines finally?

[ZcanKal]

*cough* *cough* FX Returns *cough*

[ARikozuM]

Feels like the Intel CEO has been listening to a lot of Simon & Garfunkel. 

 

 

 

[GoldenLag]

28 minutes ago, CarlBar said:

Yeah, i pointed this out a few times now but in theory given everything AMD has told us if they were willing to settle for 9900k like thermals they could

actually, if im not misstaken they can do a 95 watt 16 core at 3,7 ghz. then you add boost clocks andyou have a pretty nice CPU.

 

it doesnt neccesarly have to be toasty. (though they would probably do 105 watt TDP)

[CarlBar]

14 minutes ago, porina said:

It's called threadripper. They can't radically change AM4.

 

Yes and no, AM4 supports upto 4 dimms, i assume that means there's 4 dimms worth of pins on the bottom of the CPU. Thus in theory with an appropriate memory controller on the die they could do quad channel on a chipset that supports it. So long as they maintain backwards compatibility with existing chipsets in dual channel mode they could do it. Of course there may be some quirk to the AM4 socket i don't know about that prevents it.

 

But if they can do it i'd expect AMD to do so.

 

4 minutes ago, GoldenLag said:

actually, if i'm not mistaken they can do a 95 watt 16 core at 3,7 ghz. then you add boost clocks and you have a pretty nice CPU.

 

it doesn't necessarily have to be toasty. (though they would probably do 105 watt TDP)

 

Oh sure if they keep Ryzen 2700X clocks they could hit 16 chips at about 115w TDP, (2700x is 105w and the IO stuff isn't benefiting from a node shrink), but they'd probably still fall slightly behind the 9900K in lightly threaded workloads despite the Zen2 reported IPC uptick. And this 10 core would definitely give it a good hard kick.

[Crunchy Dragon]

I've still not had any cause to regret buying Ryzen.

 

Intel's shenanigans is just making me want to stay on AM4 more and more, especially if the Zen 3 rumors are true.

[CarlBar]

3 minutes ago, Crunchy Dragon said:

Intel's shenanigans is just making me want to stay on AM4 more and more, especially if the Zen 3 rumors are true.

 

Ohhh what haven't i heard.

 

Also love everyone's pony avatars. Daaaw.

[Crunchy Dragon]

3 minutes ago, CarlBar said:

 

Ohhh what haven't i heard.

 

Also love everyone's pony avatars. Daaaw.

7nm is supposed to be glorious, there was another Tech News thread about it detailing the rumors. Needless to say, I'm excited for it

[Canada EH]

It is well known that Intel is holding back the advancement of their product, based on whatever Ryzen is putting out.

 

Optimization of profit makes sense!

[sazrocks]

21 minutes ago, Crunchy Dragon said:

I've still not had any cause to regret buying Ryzen.

 

Intel's shenanigans is just making me want to stay on AM4 more and more, especially if the Zen 3 rumors are true.

Zen 2 or Zen 3? I haven’t heard anything about Zen 3

[manikyath]

1 hour ago, Arika S said:

The only thing worse than Intel's node size issues is their product code names

maybe its a reference to the crater it'll leave behind when cheaper VRMs explode from it's power draw.

[Arika]

1 minute ago, manikyath said:

maybe its a reference to the crater it'll leave behind when cheaper VRMs explode from it's power draw.

i was thinking "because it's will be just as hot as one entering the earth's atmosphere"

[manikyath]

3 minutes ago, Arika S said:

i was thinking "because it's will be just as hot as one entering the earth's atmosphere"

maybe it'll come on a socket the size of a small airfield, and noctua will make a cooler specificly for it, constructed with ablator heat shields.

[porina]

28 minutes ago, CarlBar said:

Yes and no, AM4 supports upto 4 dimms, i assume that means there's 4 dimms worth of pins on the bottom of the CPU.

 A channel can address multiple modules, but it can only talk to one at a time as there's shared connections.

[GoldenLag]

21 minutes ago, Arokhantos said:

I bet it runs up to 100c seeing the current trend, think i'll settle for the next thread ripper instead

Core i9 9990 10core shipping with intel stock cooler.......

[GoldenLag]

36 minutes ago, CarlBar said:

And this 10 core would definitely give it a good hard kick.

This 10 core is going to be fast, expencive and great in a multitude of ways. The only thing holding it back will be price.

38 minutes ago, CarlBar said:

(..)but they'd probably still fall slightly behind the 9900K in lightly threaded workloads despite the Zen2 reported IPC uptick

We have yet to see what boost clocks will be for these chips, but id imagine they will be high seeing as single and double core boost clocks really dont care about core count.

[GoldenLag]

15 minutes ago, Arika S said:

i was thinking "because it's will be just as hot as one entering the earth's atmosphere"

Nah its because "the heatsinks will boil off as the "sun" heats it up"

[Selah]

6 hours ago, Pascal... said:

Wow, the pike really made a hard 180 degree turn and now Intel will probably land hard on their faces for the next couple of years, lol.

in terms of consumer cpu's, yeah maybe; Intels embedded and low level tech is more than enough on its own to keep intel afloat.

[CarlBar]

19 minutes ago, porina said:

 A channel can address multiple modules, but it can only talk to one at a time as there's shared connections.

 

Ahhh, i assumed the shared stuff was on the MB side rather than the CPU socket side.

 

14 minutes ago, GoldenLag said:

This 10 core is going to be fast, expencive and great in a multitude of ways. The only thing holding it back will be price.

We have yet to see what boost clocks will be for these chips, but id imagine they will be high seeing as single and double core boost clocks really dont care about core count.

 

Based on leaks 5.2Ghz is the peak but the thing is in general we don;t see that big a gap between base and boost clocks.

[mr moose]

Why do people get hung up on price when talking about top end stuff (especially stuff most of us won't actually buy)?   The best of the best of the best always has a high price tag regardless of why.  It doesn't cost $100 more per bottle to make 12 yo whiskey but they can charge it that way because it objectively tastes better.  Same with CPU's GPU's, watches, cars etc, the difference in R+D and manufacturing might be more but they are selling to significantly less people (basically only those who want the best and don't care about cost).  The rest of us are happy to sacrifice a few performance units in order save $$.

[CarlBar]

2 minutes ago, mr moose said:

Why do people get hung up on price when talking about top end stuff (especially stuff most of us won't actually buy)?   The best of the best of the best always has a high price tag regardless of why.  It doesn't cost $100 more per bottle to make 12 yo whiskey but they can charge it that way because it objectively tastes better.  Same with CPU's GPU's, watches, cars etc, the difference in R+D and manufacturing might be more but they are selling to significantly less people (basically only those who want the best and don't care about cost).  The rest of us are happy to sacrifice a few performance units in order save $$.

 

Any uptick in peak performance usually comes with lower price point release tat uptick too. Zen2 is really going to make it's meat and bones there. AMD could release an 8 core Ryzen 5 3600X clocked at roughly the same values as a 9900K but at the Ryzen 5 price point. Given the IPC improvements on Zen 2 that should smoke everything but the 9900K and be very close to the 9900K, all whilst being less than half the price. And given what they can do with the Ryzen 7 without going about 115w TDP or so it's not remotely unreasonable they'd do that.

 

I figure the actual Ryzen 7 will be 12 cores clocked up to around the same clocks as a 9900K. Or they could do 10 cores at a slightly higher frequency.

[Crunchy Dragon]

1 hour ago, sazrocks said:

Zen 2 or Zen 3? I haven’t heard anything about Zen 3

3, I think. Whichever one was going to be 7nm.