ASUS Z170-DELUXE (Intel LGA-1151)

[cadaveca]

We've seen a GAMING board, but what about those of us who want the best while making use of all the features provided by a platform? With support for NMVe devices right in the box, the ASUS Z170-DELUXE hits the ground running. Mixed with great audio and one of the most extensive BIOSes ever, the ASUS Z170-DELUXE is exactly that - deluxe.

Show full review

 

[erixx]

I just finished upgrading 4 PCs at home (for kids and moms stuff, low-mid levels) and you are making me sick sick sick, Dave, so thanks for offering a brilliant review with totally sexy photos!!! And lots of insight that awaits further investigation My first Asus was a Deluxe, so I am droooling badly

 

[Overclocker_2001]

well this board is very good!
except sata speed, rear port ( 4 usb 3.1 will be enough... 2 usb 2.0 and 2 3.0 will be perfect! ) and.. the shroud surrounding rear port!!!
hate a lot that kind in every mobo.. seems that vrm will be sky-high temp if not cooled with a tiny hole in the rear plate...
neither p4p800-e nor p5q deluxe suffered anytime of high vrm temp so.. why this mobo ( ok heatsink is not made by copper... but is way bigger ) could need that " shroud " ?

btw interesting mobo... wifi apart :-D

 

[mouacyk]

The first section I skipped to is the OC section, brimming with enthusiasm, but was thrown off by the following:

I did do quite a bit of testing with those clocks afterward, and they proved to be quite stable, but how stable the end result is may vary from CPU to CPU.

Seeing 4.6GHz -- that's cool, about average. Seeing 1.3v -- great, this is lower than what earlier reviews have reported. Reports were up to 1.4v for this clock. However, that sentence doesn't instill any confidence into the achieved OC. Can you elaborate any more on the duration and test(s) that you used? Also (nitpicking) stability "shouldn't" vary from CPU to CPU, if you have a standard of measure. It is the OC ceiling that varies.

 

[cadaveca]

The first section I skipped to is the OC section, brimming with enthusiasm, but was thrown off by the following:
Seeing 4.6GHz -- that's cool, about average. Seeing 1.3v -- great, this is lower than what earlier reviews have reported. Reports were up to 1.4v for this clock. However, that sentence doesn't instill any confidence into the achieved OC. Can you elaborate any more on the duration and test(s) that you used? Also (nitpicking) stability "shouldn't" vary from CPU to CPU, if you have a standard of measure. It is the OC ceiling that varies.

I have an above-average 6700K, yes. I have an even better 2nd one. The one used in my board testing has a default voltage of 1.248V. My memory testing CPU has a default of 1.212V.

That sentence was referring to the stability offered by the tool's testing, rather than the CPU clock itself. Some CPUs will be close to the edge, some will not be, since the tool tests in 100 MHz increments with default settings. You are correct that the "ceiling" of some CPUs will be lower than others, and hence that comment. You simply didn't take it in stride, but no big deal. Many users complain about the auto-clocking offered by AI Suite in the past, this tool, with proper settings applied, is much better than past versions. Likely 4.7 was stable..it crashed on 4.8, and then dropped 2 multis. At that point in the review, I am reviewing the auto-clocking abilities of AI suite only.

I do tend to test far more extensively than anyone else has ever reported. Some users use a couple of tools... I use EVERYTHING. All the benchmarks shown in the review, AIDA64 stability testing (both with FPU only and with CPU/Mem/etc), Prime95, F@H, ASUS RealBench, etc, etc, etc.

I also test with a wide variety of memory modules(from 2133 MHz to 3600 MHz currently), and with more than one CPU. More than most reviewers and users for sure. So much so that your doubts in my statements are truly upsetting to me. If I wasn't honest or confident in my methods and results, I'd not even do reviews.

well this board is very good!
except sata speed, rear port ( 4 usb 3.1 will be enough... 2 usb 2.0 and 2 3.0 will be perfect! ) and.. the shroud surrounding rear port!!!
hate a lot that kind in every mobo.. seems that vrm will be sky-high temp if not cooled with a tiny hole in the rear plate...
neither p4p800-e nor p5q deluxe suffered anytime of high vrm temp so.. why this mobo ( ok heatsink is not made by copper... but is way bigger ) could need that " shroud " ?

btw interesting mobo... wifi apart :-D

VRM heat is not trapped in the shroud. Never seen it be an issue on any board I have reviewed, but if it was, I'd report about it. Those are the sorts of things I look for (see my MSI Z170 GAMING M7 review mentioning I poke around with my fingers)

 

[tabascosauz]

Would you consider adding a look at the CPU VRM to your reviews? I've seen you do it for some boards in the past, but it's on/off; I know you have a pretty good camera with which you take pictures of other surface mounted components, so would it be possible to do a picture or two of the PWM and MOSFETs? That is, on boards where the MOSFETs are reasonably easy to access; I know that some have really crappy pushpins.

 

[mouacyk]

I have an above-average 6700K, yes. I have an even better 2nd one. The one used in my board testing has a default voltage of 1.248V. My memory testing CPU has a default of 1.212V.

That sentence was referring to the stability offered by the tool's testing, rather than the CPU clock itself. Some CPUs will be close to the edge, some will not be, since the tool tests in 100 MHz increments with default settings. You are correct that the "ceiling" of some CPUs will be lower than others, and hence that comment. You simply didn't take it in stride, but no big deal. Many users complain about the auto-clocking offered by AI Suite in the past, this tool, with proper settings applied, is much better than past versions. Likely 4.7 was stable..it crashed on 4.8, and then dropped 2 multis. At that point in the review, I am reviewing the auto-clocking abilities of AI suite only.

I do tend to test far more extensively than anyone else has ever reported. Some users use a couple of tools... I use EVERYTHING. All the benchmarks shown in the review, AIDA64 stability testing (both with FPU only and with CPU/Mem/etc), Prime95, F@H, ASUS RealBench, etc, etc, etc.

I also test with a wide variety of memory modules(from 2133 MHz to 3600 MHz currently), and with more than one CPU. More than most reviewers and users for sure. So much so that your doubts in my statements are truly upsetting to me. If I wasn't honest or confident in my methods and results, I'd not even do reviews.


VRM heat is not trapped in the shroud. Never seen it be an issue on any board I have reviewed, but if it was, I'd report about it. Those are the sorts of things I look for (see my MSI Z170 GAMING M7 review mentioning I poke around with my fingers)

Sorry if I put too much emphasis on the CPU while this is technically a motherboard review, and an ASUS one at that. However, I am missing something because I can't tell if the CPU Performance results are from the automated overclock or stock. My intuition would say stock, therefore I assume there are no published OC benchmark results. In fact you didn't claim any manual overclock anywhere, so I am looking for something that doesn't exist. Well... perhaps you can manually push the chip on that board and see what you end up with, please?

 

[cadaveca]

Sorry if I put too much emphasis on the CPU while this is technically a motherboard review, and an ASUS one at that. However, I am missing something because I can't tell if the CPU Performance results are from the automated overclock or stock. My intuition would say stock, therefore I assume there are no published OC benchmark results. In fact you didn't claim any manual overclock anywhere, so I am looking for something that doesn't exist. Well... perhaps you can manually push the chip on that board and see what you end up with, please?

As you said, the end results would be CPU dependent, so are not reported.

And yes, those are stock numbers. Just like GPU reviews. Maybe I'll include one OC result in the future, like W1zzard does in his GPU reviews.

There are only minor differences in CPU OC so far in my testing, while memory OC is a bit varied by board, but only due to BIOS. So spending time reporting such things only plays into brand favoritism, in my books. If you want a board for OC, each OEM makes a board specifically designed for that, and this is not one of those boards. Proper CPU clocking on all aspects of the CPU does require a BIOS tuned to do so. When reviewing a board tuned for OC, I will focus more on OC.

I won't report on individual CPU clocking characteristics as we have a CPU reviewer to do that. I understand you might be interested to hear about my chip because it is better than most others seen in reviews, so I can tell you the max under 1.42V that I can get is 4.9 GHz, and it doesn't matter which board I use, so far. I have tested two boards fully, and the third, from Gigabyte, is currently in testing.

Would you consider adding a look at the CPU VRM to your reviews? I've seen you do it for some boards in the past, but it's on/off; I know you have a pretty good camera with which you take pictures of other surface mounted components, so would it be possible to do a picture or two of the PWM and MOSFETs? That is, on boards where the MOSFETs are reasonably easy to access; I know that some have really crappy pushpins.

PWM = Pulse Width Modulation. This refers to how the power is generated.

VRM = Voltage Regulation Module. This refers to power module components such as MOSFETs, input drivers, and chokes, as a whole.

I have a cheap $130 camera, actually (and we are talking $130 5 years ago). Good pics are about lighting and knowing the glass of the lens and how to make the most of it.

Anyway, I wonder why you care about MOSFET choices? I need a real reason to spend time investigating further, since I don't make any money doing reviews, and I know that StevenB here on TPU will include those pictures on his review on the site he reviews for. I stopped showing such when I had done enough testing to convince myself that it doesn't matter much. There are too many other variables that affect things to make providing such info worth anything other than to have pictures, and I have more pictures than most other sites already. Time is money, after all.

 

[tabascosauz]

PWM = Pulse Width Modulation. This refers to how the power is generated.

VRM = Voltage Regulation Module. This refers to power module components such as MOSFETs, input drivers, and chokes, as a whole.

I have a cheap $130 camera, actually (and we are talking $130 5 years ago). Good pics are about lighting and knowing the glass of the lens and how to make the most of it.

Anyway, I wonder why you care about MOSFET choices? I need a real reason to spend time investigating further, since I don't make any money doing reviews, and I know that StevenB here on TPU will include those pictures on his review on the site he reviews for. I stopped showing such when I had done enough testing to convince myself that it doesn't matter much. There are too many other variables that affect things to make providing such info worth anything other than to have pictures, and I have more pictures than most other sites already. Time is money, after all.

Sorry. Incorrect use of terminology. I meant pictures of the MOSFETs and the controller. Well, do what you do best. Good review as always.

 

[cadaveca]

Sorry. Incorrect use of terminology. I meant pictures of the MOSFETs and the controller. Well, do what you do best. Good review as always.

No worries; a common mistake I used to make myself all the time (even in some of my earlier reviews!)

 

[RejZoR]

ASUS finally learned to make nice software. P5Q Deluxe and current Rampage II Gene are still the traditional asian software mess with crappy bitmap skinned programs. But the new software is really nice. Sleek, easy to navigate, functional. I really like it.

 

[buggalugs]

Cool review,...... Hey cadaveca can you run a Aida memory test with the CPU at 4.4Ghz and memory at 2133?? I really want to get a clock for clock comparison between this platform and a my 4790K @4.4 GHz and memory at the same speed. I have the Z97 platform so you dont need to run that, just on the Z170 platform, if you could, it will only take a few minutes, if you're not too busy, I know you cant test for everything but if you can.

BTW, The CPU voltage setting to use on new Asus boards is Adaptive mode. Its the best setting. It allows you to run normal voltage at normal clocks and only increases voltage in turbo mode and you can set an upper limit. So, even if you set an overclock of say 4.5Ghz, with say 1.28 Volts you will only get increased voltage (1.28v) when the CPU is running between , 4Ghz-4.5Ghz, (on a 6700K) the rest of the time it runs stock voltage, including when the CPU is downclocked at idle. . With offset mode, or manual setting you're always running higher than stock voltage even if the CPU is at idle.

 

[cadaveca]

Cool review,...... Hey cadaveca can you run a Aida memory test with the CPU at 4.4Ghz and memory at 2133?? I really want to get a clock for clock comparison between this platform and a my 4790K @4.4 GHz and memory at the same speed. I have the Z97 platform so you dont need to run that, just on the Z170 platform, if you could, it will only take a few minutes, if you're not too busy, I know you cant test for everything but if you can.

BTW, The CPU voltage setting to use on new Asus boards is Adaptive mode. Its the best setting. It allows you to run normal voltage at normal clocks and only increases voltage in turbo mode and you can set an upper limit. So, even if you set an overclock of say 4.5Ghz, with say 1.28 Volts you will only get increased voltage (1.28v) when the CPU is running between , 4Ghz-4.5Ghz, (on a 6700K) the rest of the time it runs stock voltage, including when the CPU is downclocked at idle. . With offset mode, or manual setting you're always running higher than stock voltage even if the CPU is at idle.

Nope, sorry, board is already pulled apart and put back in its box and new board is being tested. You'll have to ask our CPU reviewer for CPU-specific stuff, as 2133 MHz DD4 on skylake with stock CPU (4.2 GHz) on two different boards is already posted in this review. Increasing CPU speed 200 MHz makes for only small differences (about 100 MB/s).

Also, "adaptive mode" has been around for a while now.... for several generations... and not just on ASUS, but like every brand? You'll find a recent thread with Z87 adaptive mode being talked about HERE

 

[buggalugs]

Ok, I guess I can set my CPU to 4.2Ghz to get an exact comparison.

Yah I know Adaptive mode has been around for little while , I meant Asus newer boards, I have it on my Z97, been using it since I got it , but Adaptive mode was not present on the Z87 mainstream platform, offset mode was, but not adaptive mode. It might have been present on the ROG Z87 boards but Adaptive mode first came out on Z97 for the mainstream platform. I didnt know other brands have that setting, i'm only using Asus boards lately. I just mentioned it because it sounded like in the review you werent aware of it, or preferred a different setting when overclocking. The Asus reps like AJ and most people recommend adaptive mode for overclocking these days.

https://vip.asus.com/forum/view.asp..._id=1&model=Z97-DELUXE&page=1&SLanguage=en-us

 

[cadaveca]

like I said, 4.2 to 4.4 makes for a ~100 MB/s difference. The change is very small because the L3 is separate. Increasing cache speed is the way to increase bandwidth, but it is not as "reactive" to changes as Haswell-E was, since default "cache" is 4.1 GHz. 4.2 GHz is stock on 6700K, btw, if you weren't already aware.

AH, you might be right about the mainstream platform not having it. And yeah, other brands do as well. IT's great for getting higher clocks while also maintaining power savings, but on past platforms, manual mode worked best for highest clocks, while adaptive is more suited to stock to about 4.6/4.7, depending on CPU quality.

 

[Frick]

What is the difference between the NICs? The only thing about the I219v i find is when it's about motherboards. It's not important, I just got curious.

 

[cadaveca]

What is the difference between the NICs? The only thing about the I219v i find is when it's about motherboards. It's not important, I just got curious.

You know, honestly, I'd have to look up a spec sheet. I do know that there is negligible difference between the two for most uses. I'm working on adding in LAN testing to my reviews to help differentiate between controllers, but I need to still find the "right" workload that shows real differences, but with most boards now coming with CFOS software (or KIllerNIc's similar software), managing all that and sussing out what's really going on seems a bit more complex than I had ever anticipated. Like, I suppose I could just remove the softwares... but since the board comes with software, I should use it in testing, right?

I do have a 10 Gbit system that I have set up for testing LAN devices to ensure that I can push more than most controllers, but I need to work more on tests and results before I can come up with something tangible.

 

[Joss]

Installing Windows7 is a pain, too. But if you are using the most recent hardware, you should use the most recent OS as well, and Windows 10 works great.

Why should I Dave?
If Win 7 can (and does) work fine with Z170 why make things difficult during installation? To push you to Win 10?
Well Dave, Win 10 is such a privacy nightmare that even a hardened pessimistic like myself is surprised. And a tech writer like yourself should be stressing that at every opportunity, instead of stating banalities like the one bolded above.

 

[cadaveca]

Why should I Dave?
If Win 7 can (and does) work fine with Z170 why make things difficult during installation? To push you to Win 10?
Well Dave, Win 10 is such a privacy nightmare that even a hardened pessimistic like myself is surprised. And a tech writer like yourself should be stressing that at every opportunity, instead of stating banalities like the one bolded above.

I don't feel the same as you do about Win10 and if you are worried about apparent privacy, get rid of your cell phone, and stay off the internet. I have never believed that there is any privacy on the internet, and I understand that my privacy is protected by law in my country. I use Win10 on all my machines except for one that runs Windows 8.1. I have a couple of software that I use that do not work right in Windows 10.

At the same time, it is not ASUS' fault that the 6-year old Win7 doesn't have USB drivers on it for this newly-released chipset. That's all it is. It was the same with mainstream ASUS X99 boards, too.


And no, I do not use or own a cell phone, or "personal tracking device" as I like to call it.


Also, using Windows 10 is optional, and you can slipstream driver in no problem, use Linux or even Windows 8.1. Would it have been nicer if ASUS did as MSI has, and put a work-around in the BIOS? You bet. I have pointed out that there is this problem with installing Windows7, and my obligations are complete.

 

[sonnendieb]

Awesome and thorough review. One question though: Can you use the water pump head for fans as well?

 

[cadaveca]

The water-pump plug has slightly different wiring compared to the others. It is best to check the online version of the manual to see what I mean.

 

[sonnendieb]

Hm. Not really. It's basically the same, ground, voltage, rpm and pvm. I wonder if the only difference is that the plug is full speed by default. I'm curious because I don't want to use the plugs that are somewhere in the middle of the board. Who put them there anyway? But I still Nee one more since there are five fans in my case...

 

[sonnendieb]

Hm. Not really. It's basically the same, ground, voltage, rpm and pvm. I wonder if the only difference is that the plug is full speed by default. I'm curious because I don't want to use the plugs that are somewhere in the middle of the board. Who put them there anyway? But I still need one more since there are five fans in my case...

 

[cadaveca]

Hm. Not really. It's basically the same, ground, voltage, rpm and pvm. I wonder if the only difference is that the plug is full speed by default. I'm curious because I don't want to use the plugs that are somewhere in the middle of the board. Who put them there anyway? But I still Nee one more since there are five fans in my case...

The speed is variable, but yes I do believe it is full-speed by default. In my manual the plug is wired very differently, hence my suggestion to check the online onne.

You can add more fan control by purchasing the add-on ASUS fan EXT card.

 

[CjStaal]

What extra pins are you speaking about?