An Analysis of the Z390 Socket Proves the Extra Pins Aren't Necessary

VIC-20

[H]ard|Gawd
Joined
Mar 24, 2006
Messages
1,036
Agree completely.

My favorite theory is that this new version of the socket was intended (and is still intended) to deal with the demands of the 10nm CPUs, when and if they get here.
But I won't say the pins aren't needed for the i9-9900K, there's just no proof of that.
That sounds like a very reasonable explanation to me. That being said, I fully expect a new chipset to be launched with 10nm, even if this new socket and Z390 could technically work.

Its just what Intel does.
 

Dan_D

Extremely [H]
Joined
Feb 9, 2002
Messages
56,228
Just found some Asus Z390 boards have a hidden vcore offset.... so your golden 9900K might be just average XD

I'm not sure what the reason is, but I'll clue you guys in on something. Motherboard manufacturers have been hiding settings and certain behaviors from us in the UEFI and BIOS before that for well over a decade. There are a lot of reasons for this. Some of them make sense while others are actually a bit shady. We've seen motherboard manufacturers run their front side bus, hypertransport, and base clocks slightly above Intel or AMD spec simply to win benchmarks. We've seen RAM get overvolted slightly in order to improve compatibility with certain modules. Some vendors have even offset the DRAM timings slightly to win benchmarks which usually yields to boards being more fickle about RAM than they should be.

Other settings are hidden by the manufacturer because those settings would prevent the system from running normally were you to mess with them. Sometimes a higher end and lower end motherboard actually share the same BIOS, with the lower end one simply having higher end values hidden from view or disabled entirely. All that said, I can only speculate why ASUS is doing this. However, if I were a betting man, I'd bet they were doing it to maks the inefficiency of their VRM design on many of their current motherboards. Motherboards like the Z390-A Prime, Maximus XI Hero and Maximus XI Formula all use VRM designs that are cheap to produce, but not terribly efficient.

That sounds like a very reasonable explanation to me. That being said, I fully expect a new chipset to be launched with 10nm, even if this new socket and Z390 could technically work.

Its just what Intel does.
Not necessarily. Intel has offered refreshed CPU's for the same socket many times. We saw two generations of CPU's for X58, X79 and X99 motherboards. We've seen kind of the same thing with X299. Granted, this isn't always the case with the mainstream chipsets and CPU families, but they could stick with Z390 for a theoretical 10nm CPU.
 
Last edited:

VIC-20

[H]ard|Gawd
Joined
Mar 24, 2006
Messages
1,036
Not necessarily. Intel has offered refreshed CPU's for the same socket many times. We saw two generations of CPU's for X58, X79 and X99 motherboards. We've seen kind of the same thing with X299. Granted, this isn't always the case with the mainstream chipsets and CPU families, but they could stick with Z390 for a theoretical 10nm CPU.
Yeah, its been done for sure. 945 chipset on S775 was used for Prescott and Pentium D then some could upgraded to Core 2 after a bios update. But considering how closely Z170, Z270 and Z390 are related with little or no cross compatibility, I'm not holding my breath.
 

Dan_D

Extremely [H]
Joined
Feb 9, 2002
Messages
56,228
Yeah, its been done for sure. 945 chipset on S775 was used for Prescott and Pentium D then some could upgraded to Core 2 after a bios update. But considering how closely Z170, Z270 and Z390 are related with little or no cross compatibility, I'm not holding my breath.
Fair enough. Like I said, its not generally done outside of HEDT.
 
Top