• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

ASUS Confirms Z270 Platform Could be Compatible with Intel Coffee Lake CPUs

Wasn't that because of faulty sockets?
 
It isn't about the amount of VRMs. I only mentioned the VRMs to show that ASUS really isn't the one we should be listening to when it comes to their judgement on what is necessary for power delivery.

The issue with Coffee Lake is the number of power pins in the socket itself. There aren't enough in the old 1151 socket to safely provide the power needed for the 6-core Coffee Lake. Does no one remember the burnt pins/pads problem socket 1156 had? Intel learned from their mistake and is making sure there are plenty of pins for power delivery. Yes, you could probably technically get away with the lower number of pins on the old 1151, but when you pass more current through something it gets hotter. Possibly hot enough to start melting/burning things.

When you pass too much currant through too few of pins, this happens:
burnsocketS.jpg

burnedpadS.jpg

Those images are for the faulty LGA1156 socket made by Foxconn.

OC'ed KBL and SKL already draw way more power than the standard spec for LGA1151 yet you don't hear anyone moaning about fried CPU sockets in various forums.
 
Those images are for the faulty LGA1156 socket made by Foxconn.

OC'ed KBL and SKL already draw way more power than the standard spec for LGA1151 yet you don't hear anyone moaning about fried CPU sockets in various forums.

It is still the result of too much currant through too few pins, that was my point.

Yes, Kaby and Skylake overclocked pulled a lot of current, but Coffee Lake overclocked will pull more. So more pins are needed to keep things safe.

I've been saying that the original 1151 wasn't setup for more than 4 cores since it first came out and people started asking if it would have 6 core CPUs.

Intel can't just put the product out based on it probably being good enough. They have to be completely sure the socket and processor won't be damaged by the higher power draw.
 
Last edited:
Current.

Though It could be black currant jelly stains.


So either those pins are never any good for overclocking and overvolting at any level with older processors, or.... they would be fine with the minor power increase at stock clocks for a new chip.

See how that works, if it were such an issue at the socket, any attempt to overclock and or overvolt would be met with immediate destruction. But it's not.
 
It is still the result of too much currant through too few pins, that was my point.

Yes, Kaby and Skylake overclocked pulled a lot of current, but Coffee Lake overclocked will pull more. So more pins are needed to keep things safe.

I've been saying that the original 1151 wasn't setup for more than 4 cores since it first came out and people started asking if it would have 6 core CPUs.

Intel can't just put the product out based on it probably being good enough. They have to be completely sure the socket and processor won't be damaged by the higher power draw.
Well they didn't release the 4c CL parts either for that matter on z170/z270. Intel is FoS a lot of the time on things it claims and says.
 
I can confirm that i once made an Intel Desktop Board DZ87KLT-75K compatible with the Devil's Canyon CPUs, Intel does not officially support them on Z87 Chipset, all i did was updated the Intel Management Engine and the BIOS also needs to already be detecting the CPU for this to work or if not the BIOS itself also needs to be updated

You could check the thread on intels community forum, where i posted the method i used

https://communities.intel.com/thread/78161
My ASUS Gryphon Z87 supports i7-4790K
 
So basically intel just trying to milk customers to max.

Not that Intel customers really care. The instant Coffee lake released Intel fanboy's come out praising it's performance and price. It's a good product only because AMD forced Intel to do it.
 
So either those pins are never any good for overclocking and overvolting at any level with older processors, or.... they would be fine with the minor power increase at stock clocks for a new chip.

See how that works, if it were such an issue at the socket, any attempt to overclock and or overvolt would be met with immediate destruction. But it's not.

No, that isn't how it works. Skylake and KabyLake overclocking didn't increase the currant through the socket that much. First of all, the overclocking on those chips wasn't exactly stellar. You already had parts like the 7700K running at 4.5GHz under load, and because of the thermal constraints pushing them to 5GHz, even under water, was easier said than done. That means the actual currant increase for the previous chips when overclocking was actually pretty low. When Anandtech did their 7700K reviews, they pushed the chip to 1.4v, and the power draw only increased by about 30w over the stock. The perspective here is the 8700K at stock is already at 30w more power draw than the 7700K, and when you start overclocking 6-cores instead of 4, the power draw goes up much quicker.

So, it comes down to, yes Intel probably could have made Coffee Lake work on the old socket 1151. However, it could have also started causing problems too, especially once people started overclocking them, so erroring on the side of caution is best from a business stand point.

Well they didn't release the 4c CL parts either for that matter on z170/z270. Intel is FoS a lot of the time on things it claims and says.

Yeah, they learned not to make that mistake again after the socket 775 fiasco...

And frankly, there is no reason to release 4C Coffee Lake on the old chipsets/socket. If you want 4C Coffee Lake, just go buy an i5-7600K. There's really no difference.
 
Disgusting. Z270 motherboards were not released that long ago, and now everyone who bought them are robed!!! I could say the same for Z170 motherboards - even those should support Coffee Lake chips, since the same 14 nm litography process is used since Broadwell architecture... There are no IPC improvements! The only difference from Broadwell era is the massive improvement in memory bandwidth due to DDR4 and higher clock speeds! What a pathetic excuse for a company - Intel!

This even beats platform update for Sandy Bridge chips - remember how everyone was pissed of that you needed a new 1155/2011 motherboard for Sandy Bridge chips? Everyone was bitching that Intel should have sticked with 1156/1366, but now it's a standard thing that you need to change a motherboard every 2 generations - NO SIR, now you have do this every generation!

What happened to S775 days - so much architecture changes on single socket!!! In the future we will be needing to buy a new motherboard for the same family processors!

I fell pity for all Z170/Z270 buyers.
 
No, that isn't how it works. Skylake and KabyLake overclocking didn't increase the currant through the socket that much. First of all, the overclocking on those chips wasn't exactly stellar. You already had parts like the 7700K running at 4.5GHz under load, and because of the thermal constraints pushing them to 5GHz, even under water, was easier said than done. That means the actual currant increase for the previous chips when overclocking was actually pretty low. When Anandtech did their 7700K reviews, they pushed the chip to 1.4v, and the power draw only increased by about 30w over the stock. The perspective here is the 8700K at stock is already at 30w more power draw than the 7700K, and when you start overclocking 6-cores instead of 4, the power draw goes up much quicker.

So, it comes down to, yes Intel probably could have made Coffee Lake work on the old socket 1151. However, it could have also started causing problems too, especially once people started overclocking them, so erroring on the side of caution is best from a business stand point.



Yeah, they learned not to make that mistake again after the socket 775 fiasco...

And frankly, there is no reason to release 4C Coffee Lake on the old chipsets/socket. If you want 4C Coffee Lake, just go buy an i5-7600K. There's really no difference.
Less cache...actually getting a Skylake based Xeon with BCLK is probably arguably better if you intended to OC since they had 8MB cache...and yes there is point more cache helps plus if what they say is true about the refinements to 14nm that is better as well. Intel is just milking consumers they could have released the i3 coffee lake chips on older chip sets and newer ones on the newer z370 and saved consumers the cost of a unneeded motherboard upgrade. Also the older Skylake/Kaby Lake chips have gone up in price rather than down over time which is the opposite of what you'd expect for a older supposedly "less powerful" chip. Intel has created a artificial supply and constraint with the way it keeps changing chip sets and CPU skew's and phasing others out prematurely to push the demand for the newer chip sets and CPU's up further. I don't remember much fiasco with LGA775 other than P4's running hotter than a stuck on brake caliper. Hell LGA775 even supports DDR3 2200MHz on a couple P45 boards probably higher technically it's just DDR3 was in it's very early infancy and that's w/o fancy north bridge chip set cooling would helps a lot in that regard.
 
Less cache...

Sure, but the extra 2MB of cache on the 8350K doesn't really help it any. The 7600K performs nearly identical. And the 7700K is still better than the 8350K as well, so people with the older platforms have options that are as good or better than the 8350K.

plus if what they say is true about the refinements to 14nm that is better as well.

They've been on 14nm for 4 generations now. I really don't think there is much more they can do to refine it between the 3rd and 4th generation of chips.

Intel is just milking consumers they could have released the i3 coffee lake chips on older chip sets and newer ones on the newer z370 and saved consumers the cost of a unneeded motherboard upgrade.

Instead they'd just be getting unneeded CPU upgrades. There is no point to release a 4C Coffee Lake on a platform that already has a 4C Kaby. There really is no worthwhile improvement. And releasing half a product line that works on older motherboards and half that doesn't leads to consumer confusion.

I don't remember much fiasco with LGA775 other than P4's running hotter than a stuck on brake caliper. Hell LGA775 even supports DDR3 2200MHz on a couple P45 boards probably higher technically it's just DDR3 was in it's very early infancy and that's w/o fancy north bridge chip set cooling would helps a lot in that regard.

Then you should brush up on your computer history. They did exactly what you are asking them to do with 775, and it backfired. People were buying computers and motherboard that worked with Core 2 Duos just fine, but when they went to upgrade to Core 2 Quads, they found their motherboards didn't support them. Chips from the same product line, from the same generation, wouldn't all work in the same motherboards. I don't think Intel is going to make that mistake again.
 
The 8350K is better value though than the 7700K and if the cache didn't help Intel's cheap bastards ***'s wouldn't have it they love to cut corners to save a penny. The 8350K actually has better SC and quad core OC than 7700K it does and is better at QC and SC though a bit worse at MC due to lacking hyper threading. You are inherently wrong that it would be bad to offer those things to z170/z270 motherboard owners that gap will actually widen further in terms of value because 7600K/7700K prices will continue to increase if the past is any indicator.
http://cpu.userbenchmark.com/Compare/Intel-Core-i7-7700K-vs-Intel-Core-i3-8350K/3647vs3935

At the end of the day however Ryzen is still the best value by a long shot relative to overall performance to me. Now if I didn't have to buy a new motherboard on top of CPU the 8350K could have been tempting for z170/z270 users like myself so Intel's loss AMD's gain as far as that goes. I'll just score a 6700K off ebay in another year or two once they are cheap as hell and headed to a landfill for disposal shortly otherwise priced. I'll save my money toward a GPU instead bang for buck king.
 
...and yet we have people here on this board that defend Intel everyday. Intel is one of the advancement killers of today. AdoredTV made a pretty good video on what Intel actually is (hint: Intel is just about money, nothing else):
Welcome to the world of Capitalism. Please enjoy your stay.
 
I hope amd make improvements to the new ryzen lift incoming February (especially to single core performance) cause lot owner s of older CPU - sockets have disgusting with Intel's policy
I'm using 7700k
 
The 8350K is better value though than the 7700K and if the cache didn't help Intel's cheap bastards ***'s wouldn't have it they love to cut corners to save a penny. The 8350K actually has better SC and quad core OC than 7700K it does and is better at QC and SC though a bit worse at MC due to lacking hyper threading. You are inherently wrong that it would be bad to offer those things to z170/z270 motherboard owners that gap will actually widen further in terms of value because 7600K/7700K prices will continue to increase if the past is any indicator.
http://cpu.userbenchmark.com/Compare/Intel-Core-i7-7700K-vs-Intel-Core-i3-8350K/3647vs3935


Userbench is quite possibly the worst place you can get comparative results. Go read some actual reviews. A benchmark "suite" that takes less than 2 minutes to run is not properly testing the performance of the CPU. Look here: https://tpucdn.com/reviews/Intel/Core_i3_8350K/images/perfrel_cpu.png Notice how the 7600K is virtual identical to the 8350K. I don't really think anyone with a 7600K is running out to the stores to upgrade to a 8350K, even if it was compatible with their motherboards.

Oh, and lets talk about value. The 7600K right now is cheaper than the 8350K. The 8350K is going for, what about $240 right now. The 7600K can be had for $230. Even if the 7600K does go up a little, it won't really be that much. History actually shows the processors don't really go up in price. Example, the 6600K is still available for $230 too. Plus, you're far more likely to find a good value on the 7600K in the used market than an 8350K, used 7600Ks go for $200 right now. And the used price is going to drop even more if the price of the 8350K comes down any. The way prices really work is the previous platform's processors drop in price on the used market significantly after a new platform is released and people are dumping their old parts on the market. Then as the used supply dries up, the price goes back up to about what the original retail price was. Hell, you can still buy a brand new 6700K for $315. The price hasn't gone up any, heck it's gone down. When the 6700K came out, it was $350.
 
:respect: Well you are right :laugh: Ryzen 1600/1700 are a better value :roll:

As soon as you can figure out how to put a Ryzen 1600 in a Z170/270 board, you'll have a valid point.
 
Oh, and lets talk about value. The 7600K right now is cheaper than the 8350K. The 8350K is going for, what about $240 right now.
I see 8350K in stock at 169.90€ (euro prices with taxes).
Intel 8000 series availability seems to be a lot worse on the other side of pond.

:respect: Well you are right :laugh: Ryzen 1600/1700 are a better value :roll:
1600, probably yes. On the other hand there is 8400 that currently sits right at the same price point as 1600 (with current inflated prices). 1700 is twice as expensive as 8350K, not a good comparison.
 
Ill cash in on a Z370 board easily, just cause it doesnt work for others. Then upgrade to a cheap CPU to test with.....
Best part is the resell value after repair.
 
I see 8350K in stock at 169.90€ (euro prices with taxes).
Intel 8000 series availability seems to be a lot worse on the other side of pond.

1600, probably yes. On the other hand there is 8400 that currently sits right at the same price point as 1600 (with current inflated prices). 1700 is twice as expensive as 8350K, not a good comparison.
Yeah BS 1700 isn't a good comparison a CPU + MB for Ryzen 1700 is cheaper than a 8350K + motherboard. It's also not twice as expensive it's $30's more.
 
Intel plan to launch Coffee Lake with 8 cores?
Problem is because people are too much opsessed with last models not with performance.
X99 platform is compatible with Haswell-E and Broadwell-E, almost 3 years.

Now if someone have good motherboard, DDR4 3000-3200MHz don't need to upgrade on X299 8 cores. Better to wait price drop of i7-6950X and upgrade on him.
But Intel need to slow down with new chipsets
 
Last edited:
X platform is still more attractive to me. No matter on slightly smaller fps in games.
You buy some ROG motherboard, some cheaper version of CPU, example i7-7820X and install Monoblock on Rampage 6 Apex motherboard, with one GTX1080Ti Poseidon, 10/13mm fittings, no need for bigger, X-Res Revo 100 and wait little to price of i9 drop. Don't need to be newest platform to satisfied all needs of enthusiasts and gamers. I like to keep platforms as X58, X79, X99, X299 longer than P55, P67, Z68, Z77, Z87, Z97, Z170, Z270, Z370.
4 chipsets and 2 sockets Xtreme vs 9 chipsets and 4 sockets Mainstream.
Difference from Z77 to Z370 is small for 3 sockets. OK now difference is very nice compare to Z77-Z270.
Coffee Lake i7-8700K is long waiting 6 core for mainstream. People who thought about upgrade their i7-3770K, i7-4770K no need to wait nothing else.
Intel made good move, i5-8600K best option for gamers who don't want fancy and expensive processors.
6 core without HT, he will OC better and kill i7-7700K completely.
 
Back
Top