Z170 CLAIMED to be hacked to support Coffeelake

[Super Nade]

My friend corpsejockey on here alerted me to this bit of news. I am not sure of its veracity, i.e. it could be fake.

A user recently posted on Baidu that they were able to hack their MSI Z170A XPower Titanium BIOS, allowing it to support the Intel Core i3-8350K Coffee Lake CPU.

https://www.tweaktown.com/news/6003...ake-8350k-running-z170-motherboard/index.html

I'll leave it up to use guys to check if this is corroborated.

 

[custom90gt]

Interesting, I thought it was a different pin out since intel didn't release the coffee lake specs even for z270? Hacking the bios to support CPUs usually just means injecting an updated/new CPU microcode, not sure what else they would have done to get PCI and igpu issues...

 

[mimart7]

Also related, https://thehackernews.com/2017/11/intel-chipset-flaws.html

At my workplace we are evaluating any machines that may be impacted.

 

[corpsejockey]

It appears more progress has been made. The wording gets a little confusing at times, but it would seem asrock z170 boards have been made fully functional with CL. Expirements with other brands have had some amount of success, but still have pcie and igpu issues atm.

https://www.win-raid.com/t3251f16-Coffee-lake-Cpus-on-Sky-Kaby-Lake-chipsets.html

Trick seems to revolve around cpu microcode, pcie strings, and downgrading management engine version

 

[Woomack]

ASUS officially said they could make their z270 mobos support coffee lake but because of their agreement with Intel, they won't. If Intel had stronger competition then they wouldn't make moves like that.

 

[Nebulous]

Yeah seriously, Intel is using strongarm tactics to keep us (Z270) users from having the ability to upgrade just to CPU saving us a boatload of money, and forcing us to a complete upgrade to jump to the Z370 platform. Kinda sux IMHO, but doesn't surprise me.

For this I'm reconsidering my upgrade path and looking at AMD to see if they bring out something that will be upgrade worthy and away from Intel.

 

[corpsejockey]

Yep, I also have an x370 and a r5 1500. My next move may be a ryzen 2xxx.

 

[wingman99]

I read the from one of the ASUS vendors that with the Z370 there is two CPU power pins on the 1151 socket for the increased current instead of one for the Z170, Z270.

 

[corpsejockey]

I read the from one of the ASUS vendors that with the Z370 there is two CPU power pins on the 1151 socket for the increased current instead of one for the Z170, Z270.

Yeah, and apparently they connect internally to the same power plane lol.

 

[wingman99]

Yeah, and apparently they connect internally to the same power plane lol.

It may be the same power plane from the VRM, however the tracers on the board to the 1151 socket pin out on the Z370 board are different than the Z170 and Z270 board.

 

[corpsejockey]

I think that's where the address strings come into play.

 

[wingman99]

I don't know of any address strings. There is the VRM Vcore power rail to the motherboard tracers to LGA 1151 pins.

 

[corpsejockey]

Sorry, was talking about remapping for igpu and pcie lanes and that is probably inaccurate description. Power delivery pins are different but apparently doesn't matter as the additional power pins on cl were reserved/unused in sl and kl. those pins are supposedly connected to the same power plane in the cpu regardless of what traces on the board go to what pin. Please note... I don't know anything... I'm a noob. It says so right under my name. Im going off some random broken English post that I only understand about 90% of

I have looked over the pin differences. Most reputable tech sites say it's not possible, but we had asus unofficial statement that they could have made it work. Then what appears to be confirmed report of 8350 working on z170 board all be it with crippled igpu and pcie.

Is this fake news?

 

[wingman99]

Sorry, was talking about remapping for igpu and pcie lanes and that is probably inaccurate description. Power delivery pins are different but apparently doesn't matter as the additional power pins on cl were reserved/unused in sl and kl. those pins are supposedly connected to the same power plane in the cpu regardless of what traces on the board go to what pin. Please note... I don't know anything... I'm a noob. It says so right under my name. Im going off some random broken English post that I only understand about 90% of

I have looked over the pin differences. Most reputable tech sites say it's not possible, but we had asus unofficial statement that they could have made it work. Then what appears to be confirmed report of 8350 working on z170 board all be it with crippled igpu and pcie.

Is this fake news?

It's not about working, i'm sure you could with a BIOS modification to allow the boot of coffee lake. Intel's specification for Z370 either runs dual tracers or a larger tracer from the VRM to the two LGA pins on the Z370 to double the allowable current delivery using two Vcore pins for 6 cores under full load so the two LGA pins and tracers does not overheat like it would with one pin and small or single tracer on the Z170 Z270, if you ran it that way.

Let me simplify it, some video cards have two power cables from the same PSU. You could just use one power adapter with some modification right.

 

[corpsejockey]

Lol, its cool. No need to simplify. I'm a design engineer... I size electrical circuits daily. Its much larger scale and less complicated, but same principle. It just appears that higher end boards with more robust vrm could handle cl. I didn't mean for this to be a sling mud at Intel thread. I assume the decision was made based on spec and what random mb manufactures have done on higher end boards above and beyond basic spec was irrelevant. I personally think in some part this was intel scrambling to get a consumer 6 core out the door due to amd's growing popularity rather than strictly an evil plot to get my money.

I am interested in how this pans out though. Only way I would even consider cl at this point is if I can get it working on one of the z270 boards I already have.

 

[Alaric]

this was intel scrambling to get a consumer 6 core out the door due to amd's growing popularity rather than strictly an evil plot to get my money.

Meh, same thing. LOL

 

[wingman99]

Lol, its cool. No need to simplify. I'm a design engineer... I size electrical circuits daily. Its much larger scale and less complicated, but same principle. It just appears that higher end boards with more robust vrm could handle cl. I didn't mean for this to be a sling mud at Intel thread. I assume the decision was made based on spec and what random mb manufactures have done on higher end boards above and beyond basic spec was irrelevant. I personally think in some part this was intel scrambling to get a consumer 6 core out the door due to amd's growing popularity rather than strictly an evil plot to get my money.

I am interested in how this pans out though. Only way I would even consider cl at this point is if I can get it working on one of the z270 boards I already have.

It's not about the VRM's on the Z170 Z270. There is no tracer (wire) from the VRM rail connected to the extra LGA 1151 pin Intel is using for the increased AMP's that the 6 core does draw.
Motherboards are very simple in design, that is why mine only cost $99.00, they just have PCB (Wires) with chips and everything that needs to be connected is, they don't waste money on copper connection not needed for the series.

The Z170 Z270 VRM's can supply enough AMPs however the PCB (wires) and one LGA 1151 pin would overheat (burn) with coffee lake. I have never seen a BIOS update for a processor upgrade do to motherboard manufactures not following Intel's specification previously. As a design engineer you know for the extra amps the 6 core does draw the motherboard has to be designed from the ground up.

In 2016 the Z370 was planned, next will be Z390. Intel already new how to make 6+ core power delivery for the high end desktop and sever processors.

 

[corpsejockey]

Yeah, but we are not talking about 1 pin. The amp draw was not dramatically increased and they didn't double the vcc pins. They added 18 vcc pins. Kl had 128, cl has 146. That's 13 to 14 % max amp draw increase through the pins. I admittedly have not gone through the architecture to break down what pins powers what modules in the chip and how the power is distributed. The op in the thread I linked claims to have though and stated it was only 8% improvement on power impedance.

I do strongly disagree that this requires and new mobo design from the ground up.

 

[wingman99]

Intel's maximum watts specification for coffee lake is 209w and kaby lake 152w = 37% increase in watts, the new Z370 board had to be done. Folks overclocking at overclock.net are having to add cooling for VRM's and increas power limits with high end Z370s. Overclocking is not like kaby lake.

The 14% max amp draw increase for coffee lake Z370 does not show how much buffer room the kaby lake Z270 had for total AMP draw.

 

[UltraTaco]

But 14% is a big deal? I remember reading Tom's article on i7 920 how they clocked it from 2.6 to 4ghz and power draw increased like 50% ! Don't remember if it was total from outlet or just cpu. Pretty sure it was mostly cpu..
How can you OC if mobo will burn traces at 14% amp draw increase?