Sky/kabylake in z370

[Fossburger]

Fresh off the new wan show.. with ROG saying they could be made compatible but are locked out... why do you think older 170/270 cpus would be locked out of being put in the new 370 mobos?

Especially since you basically cant get your hands on a new 8700k anyway so placing the older 6600k in there as a place holder until stock arrives for coffeelake would be nice

[ARikozuM]

There may be an issue with the power pins on the motherboard. Perhaps if they can't be turned off it would cause issues with excess power, but I would not know. 

 

@done12many2 @MageTank @HKZeroFive

[DocSwag]

Create more coffee lake sales I presume.

[DocSwag]

Just now, ARikozuM said:

There may be an issue with the power pins on the motherboard. Perhaps if they can't be turned off it would cause issues with excess power, but I would not know. 

 

@done12many2 @MageTank @HKZeroFive

I would personally guess that wouldn't be an issue. The power would just go through all the other pins instead and those few pins essentially won't be used.

[Cyracus]

pretty sure one of the tech tubers tried it, worked about as well as shoving a potato in the socket...

[ARikozuM]

2 minutes ago, DocSwag said:

I would personally guess that wouldn't be an issue. The power would just go through all the other pins instead and those few pins essentially won't be used.

Like I said, I wouldn't know. I won't say that it would hurt or not. 

 

Especially with the Intel hateboys all over the damn AO. 

[MageTank]

I personally think it's a mix of both Intel's arbitrary locking (to adhere to their usual 2 generations per chipset trend) and a ploy for future compatibility. As per usual, I cannot say how I know, but it's a fact that cannonlake is compatible with "300 series PCH". Z370 is technically a 300 series PCH. We have improved power delivery, and linus hinting that he knows why they did it, but cannot say due to an NDA of some kind. What could possibly be under NDA now, that is related to the chipset/socket configuration? Cannonlake, lol.

 

I will say this, that increased power delivery is real. Look at the memory overclocking on CFL and tell me it isn't. Even budget boards like the Z370 Pro4 has rated speeds up to 4266. This was unheard of on Skylake/Kaby unless you had boards like the OC Formula, the Apex, or the ITX enthusiast boards. 

[Fossburger]

2 minutes ago, MageTank said:

and linus hinting that he knows why they did it, but cannot say due to an NDA of some kind. What could possibly be under NDA now, that is related to the chipset/socket configuration? Cannonlake, lol.

the way i read Linus' expression was that someone had told him in confidence and Linus simply didnt to jeopardise anything.. no NDA just simply respect for his source

 

[MageTank]

Just now, Fossburger said:

the way i read Linus' expression was that someone had told him in confidence and Linus simply didnt to jeopardise anything.. no NDA just simply respect for his source

 

He implied that there was a legitimate reason, and that he knew it, and that it is not public knowledge. CFL is out, it's been launched and all of the information relative to CFL is publicly available in Intel's whitesheets. My point is, what could possibly be related to CFL and Z370, that is not public knowledge, that would be a legitimate reason for the effort Intel put into designing a new socket? 

 

As I said, Cannonlake is the next generation after CFL, and it's slated to be supported on "300 series PCH" boards. Z370, while a carbon copy of Z270 when comparing strictly chipsets, still technically fits that description. 

 

Though, I do see your point, there is a chance this isn't an NDA at all. However, my original point still stands. 

[Trulop]

Honestly, it could be as simple as, "what if its not completely stable in all scenarios?"

If they "supported" the new cpus in the older chipset boards and happened to have some kind of malfunction or have it fried while trying to overclock (in the safe parameters you could on z370) they could be held accountable when everyone points the finger at them saying, "you said it would work!" as we know alot of people would do lol. Im not saying i dont think they could have attempted to work around this because they rushed these chips out to compete with ryzen where they might have been able to find a solution, but in terms of what they had when they went to release, im putting my cards in they didnt really have a choice. If they would have been able to get all the bonuses and gains from the new chips while being able to utilize the other boards, the possibility of others wanting to upgrade either now or a bit later, could have their profits even higher that limiting it to another whole new platform change and if we know intel, they want your money  

Just my opinion on the matter, could be very different, we may never know.

[HKZeroFive]

Aside from Intel's "cuz power delivery" explanation, I'd argue that it also doesn't make any sense to put an older generation chip into a Z370 motherboard anyway.

 

However, it is a very artificial limitation. That's undeniable.

[DocSwag]

10 hours ago, MageTank said:

I will say this, that increased power delivery is real. Look at the memory overclocking on CFL and tell me it isn't. Even budget boards like the Z370 Pro4 has rated speeds up to 4266. This was unheard of on Skylake/Kaby unless you had boards like the OC Formula, the Apex, or the ITX enthusiast boards. 

I would personally think this is more likely to be from a better IMC or better trace layouts. Socket differences wouldn't really make up for this since it only affects v core and ground.

 

Really, if Intel wanted to they totally could have done it. It's mostly for segmentation and sales that they didn't.