ASUS Z97 Motherboards - Official Support Thread

Raja@ASUS

ASUS Community Support
Joined
Apr 18, 2011
Messages
2,390
This thread is for the official support of the ASUS Z97 motherboard series.

Lots of feature changes over the older platforms - from software, UEFI and aesthetics.

Motherboard Info


A comparison table containing all ASUS Z97 motherboard features can be found here: http://www.asus.com/us/site/motherboards/Z97/comparison/


Key Features For ASUS 9 Series Motherboards​



New UEFI Features



EZ_setup.png

EZ Setup Screen - One-stop location for simple system setup..



The EZ mode screen allows quick setup of storage BOOT priority, XMP for memory, RAID, overclocking presets via wizards, system time, fan control and many more features.





manual_fan_control.png





Fan Controls are now far more intuitive with a graphical interface for setting the fan slope with 3 control points (more than enough to configure any fan) in manual mode as well as 4 preset modes. Full PWM and DC support is present on EVERY fan header across all ASUS Z97 motherboards from the Z97-A to Deluxe, ROG and of course the TUF series.

We've also introduced the ability to map each fan to various temperature sensors on the board instead of having all system fans react to the CPU temp only. For example, the chassis inlet fans can be mapped to PCH temperature which allows the fans to react to temperature changes in that region - GPU environment temperatures for example.

A fan calibration routine has also been added to UEFI advanced mode. The calibration routine determines the minimum rotational speed of each fan and then sets the fan slope minimum value accordingly.

With these features, users can enjoy comprehensive fan control with any operating system as no software is required (although we still have Fan Xpert 3 in Windows - more on that later).





Major Software Changes


Graphically AI Suite 3 is the same as the Z87 series but there have been a lot of functionality refinements that take the software to the next level.




5-Way Optimization - Auto Tuning (Mainstream and WS Only)



Auto Tuning is an automated overclocking routine that tunes a system overclock for the end-user. We’ve had this feature for a few generations now and have been enhancing it year-on-year based upon end-user, media and our internal feedback.

Auto tuning works by running a Prime based stress test on the system. Previously, there was no control over the duration of the stress test. We ran a quick 15 second test and then increased processor clock frequency in steps. For Z97, we've added the ability to set stress test duration up to 1 hour at each step (we can extend this further via an update). Plus we've added a memory stress test to the mix for more stringent overclock testing.

Users can also target frequency or a temperature to tune to. Overall a far more comprehensive set of auto tuning parameters to help make overclocking easy for newcomers and help seasoned users evaluate the overclocking potential of a CPU effortlessly.





Turbo App​



Turbo App is the 5th element of 5-Way Optimization. This nifty addition allows us to set dynamic overclocking profiles based upon the application being used.


The advantage of this is two-fold:

1) Overclocks no longer need to be limited by the hottest application we run on our system. Typically, we tune our maximum overclock to the hottest software we will run. With Turbo Ap, we can set higher overclocks for software/games that don’t load the CPU as heavily.

2) The overclock is applied dynamically in real-time as we switch through applications – no need to enter UEFI to change the overclock preset or navigate via various GUIs changing configurations. Turbo App also contains LAN priority and audio presets which will change to our desired setting when the application is active.​

Turbo App is simple to use and setup and helps get the most out of each CPU.





Fan Xpert 3​

While we have added a comprehensive set of fan control features in UEFI, there’s more on offer in Fan Xpert 3.



i) Full fan calibration for each fan connected to the motherboard.

ii) The ability to control fan spin up and spin down response.

iii) Setting lower speeds for fans that can spin at low duty cycles.

iv) Look up tables for each fan to show RPM versus power or duty cycle.

v) Renaming of each fan for easy reference.​


With all of these options at our disposal, just about any fan con be configured to be quiet and react to system temperature changes. Between UEFI and Fan Xpert 3, the options on offer are the most comprehensive ever seen for motherboard fan control. In fact, there’s no reason to invest in a dedicated fan controller at all.




Push Notice



Push notice is an application that teams a mobile device such as a phone or tablet to the PC in order to send system event, sensor or user messages to the remote device. Once a device is paired to the app (requires client software – which is free to download and use) the alerts will work over a local network or via the internet. This means the messages will be received no matter where you are, provided internet access is available.




push_notice.jpg




System events such as configured restarts, system shutdown or sleep can be set and a message sent to the remote device at a user configured time before the event occurs.


push_notice1.jpg




Power supply voltage and CPU package temp warnings can also be set to trigger an alert message to the paired remote device.






Revised Audio deisgn (mainstream)​



Mainstream motherboard audio has been revised for 9 series – physically similar layout techniques to the ROG boards to prevent crosstalk.

Other changes include a RTL circuit to prevent DC thump for all outputs while the Deluxe model gets a buffer amplifier to drive either the front headphone jack or rear output jack. The buffer is only needed when connecting hedphones with an impedance over 120 ohms or so to provide additional voltage gain and prevent excessive roll-off at frequency extremes.






Product Page Links​



Z97-Deluxe : http://www.asus.com/Motherboards/Z97DELUXENFC_WLC/
Z97-PRO : http://www.asus.com/Motherboards/Z97PRO/
Z97-A : http://www.asus.com/Motherboards/Z97A/
Maximus VII Hero : http://www.asus.com/Motherboards/MAXIMUS_VII_HERO/
Maximus VII Gene : http://www.asus.com/Motherboards/MAXIMUS_VII_GENE/
Z97 Sabertooth Mk.1 :http://www.asus.com/Motherboards/SABERTOOTH_Z97_MARK_1/







Guides​


5-Way Optimization User Guide


UEFI Fan Control and Fan Xpert 3 Guide



Latest UEFI Updates:


Z97-Deluxe: https://drive.google.com/file/d/0Bz2VRRbLPrZnckx1QmNHX0xQdVU/edit?usp=sharing

Z97 Sabertooth Mk.1: https://drive.google.com/file/d/0Bz2VRRbLPrZnTmNFb3VzSTRJdnM/edit?usp=sharing

Z97-Pro: https://drive.google.com/file/d/0Bz2VRRbLPrZnQ29DR25UUmpHdlU/edit?usp=sharing

Z97-A: https://drive.google.com/file/d/0Bz2VRRbLPrZnWktvZjhnWkJvMzA/edit?usp=sharing

Max 7 Hero: https://drive.google.com/file/d/0Bz2VRRbLPrZndTFNVVpIRF9vQkE/edit?usp=sharing

Max 7 Gene: https://drive.google.com/file/d/0Bz2VRRbLPrZnSHJfMWNpNjFfT0U/edit?usp=sharing


Patch list:


1. Update uCode for new Anniversary Edition CPU support.
2. XMP DRAM frequency reset after load optimized defaults.
3. Enhance security setting.
4. Enhance compatibility for ASUS ThunderboltEX II series cards.
5. Enhance stability for OC profile function.
6. Enhance legacy storage support for the RAID wizard in EZ Tuning Wizard function.
7. Change “Next Boot after AC Power Loss” setting to speed up PC boot time.

These builds will be up on the support site shortly.
 
Last edited:
Other changes include a RTL circuit to prevent DC thump for all outputs while the Deluxe model gets a buffer amplifier to drive either the front headphone jack or rear output jack.

This feature already existed on the Max 6 ROG boards? I just got a set of Klipsch ProMedia 2.1 Wireless speakers and they thump horribly when shutting them off/on. I think it's a problem with the speaker design and has nothing to do with the board... The set of Altec Lansings I was using previously did not thump loudly when mains were shut off and now I have a master control Battery Backup that shuts down other outlets after a configurable delay when the motherboard goes into sleep or is shut down--so the speakers aren't even shutting off until the computer is on standby power. Then, big boom. Think the speakers need to be returned or left on 24/7...
 
Last edited:
I need "SATA Express for dummies"
So like..If I RAID 0 2 ssds on z97;Will it be faster than on z87?
 
This feature already existed on the Max 6 ROG boards? I just got a set of Klipsch ProMedia 2.1 Wireless speakers and they thump horribly when shutting them off/on. I think it's a problem with the speaker design and has nothing to do with the board... The set of Altec Lansings I was using previously did not thump loudly when mains were shut off and now I have a master control Battery Backup that shuts down other outlets after a configurable delay when the motherboard goes into sleep or is shut down--so the speakers aren't even shutting off until the computer is on standby power. Then, big boom. Think the speakers need to be returned or left on 24/7...

No, this is not on the ROG boards. Those boards have a relay on one set of outputs.
 
No, this is not on the ROG boards. Those boards have a relay on one set of outputs.

I get it :) I read the audio section wrong. It's off topic slightly but I reiterate that my DC thumping problem with the Klipsch ProMedias seems unrelated to the motherboard's circuits.
 
Raja when will the ROG Ranger be available in the US for purchase. Need to buy a new rig this week hope it is soon.
 
Updated first post with new UEFI builds:

Z97-Deluxe: https://drive.google.com/file/d/0Bz2VRRbLPrZnckx1QmNHX0xQdVU/edit?usp=sharing

Z97 Sabertooth Mk.1: https://drive.google.com/file/d/0Bz2VRRbLPrZnTmNFb3VzSTRJdnM/edit?usp=sharing

Z97-Pro: https://drive.google.com/file/d/0Bz2VRRbLPrZnQ29DR25UUmpHdlU/edit?usp=sharing

Z97-A: https://drive.google.com/file/d/0Bz2VRRbLPrZnWktvZjhnWkJvMzA/edit?usp=sharing


Patch list:


1. Update uCode for new Anniversary Edition CPU support.
2. XMP DRAM frequency reset after load optimized defaults.
3. Enhance security setting.
4. Enhance compatibility for ASUS ThunderboltEX II series cards.
5. Enhance stability for OC profile function.
6. Enhance legacy storage support for the RAID wizard in EZ Tuning Wizard function.
7. Change “Next Boot after AC Power Loss” setting to speed up PC boot time.

These builds will be up on the support site shortly.
 
I have a Maximus V Gene and I'm looking at the VII Gene and Gryphon. It looks like the VII Gene is out, but the Gryphon is not in stock anywhere - any idea on when it's coming?

I like the DTS encoding on the Gene but I don't really need it. I also like the cool remote features on the Gryphon and I wish you guys would not split features like that, I'd like a board that has it all.

Does the sound daughter card interfere with SLI at all? Can you speak to the differences in the Networking features? This system will be mostly for NV Surround gaming - I plan to bring over most of my system including RAM.

Thanks,
 
Got an M7Hero up and running here.

Super impressed with feature set. Triple/double cold booting no more. Software is a lot better now. Excellent fan control. Was actually able to overclock my chip with its autotune to 4.6 without flipping out. It just did it.

1 negative and a couple confusions so far:
I guess they removed a couple of the USB ports to add a DVI port... not sure that was a good move, but if they improved the stability of the USB ports themselves I can live with it. They seem physically sturdier. 6 rear USB ports is enough, I suppose, but 8 was definitely better.

Anyway, one question is about the UEFI setting for SVID support. I know for a fact on the M6H bios it says to disable SVID when overclocking. The M7H UEFI says exactly the opposite. What is SVID support and what is with the change in instructions on using it? It's been off for the entire time I've been using a 4770k.

Other question is more depressing. On the M6H you can trick the manual voltage mode into dropping voltage as if you are using adaptive... but it completely caps voltage. This was the ideal setting for me. On the M7H this trick no longer works. Unfortunately, Adaptive mode set to 1.28v regularly get 1.36v... This is a serious problem. What is going on with adaptive voltage and how can I use this thing without voltage spikes and still get low voltage while idle?
 
Hi,

1) SVID message needs to be corrected - follow M6 rules. SVID support communiccates with the external 1.8VDC supply to support phase shedding etc.

2) Adaptive is programmed by intel. When a AVX instruction set is seen it may ramp voltage beyond the set voltage - nothing at all we can do about that.. This is known by Intel. Just how they designed the CPUs. Your other alternative is to use an offset voltage, but the issue with offset is that the change to VID affects the enitre VID stack, so that can be an issue for some CPUs. Adaptive affects the Turbo portion of the VID stack only.

-Raja
 
I have a Maximus V Gene and I'm looking at the VII Gene and Gryphon. It looks like the VII Gene is out, but the Gryphon is not in stock anywhere - any idea on when it's coming?

I like the DTS encoding on the Gene but I don't really need it. I also like the cool remote features on the Gryphon and I wish you guys would not split features like that, I'd like a board that has it all.

Does the sound daughter card interfere with SLI at all? Can you speak to the differences in the Networking features? This system will be mostly for NV Surround gaming - I plan to bring over most of my system including RAM.

Thanks,

You can check the spec pages an marketing for the differences in software on the two boards. Check there, then come back and ask questions on things you don't understand. Pointless me laying out what's there already.
 
Hi,

1) SVID message needs to be corrected - follow M6 rules. SVID support communiccates with the external 1.8VDC supply to support phase shedding etc.

2) Adaptive is programmed by intel. When a AVX instruction set is seen it may ramp voltage beyond the set voltage - nothing at all we can do about that.. This is known by Intel. Just how they designed the CPUs. Your other alternative is to use an offset voltage, but the issue with offset is that the change to VID affects the enitre VID stack, so that can be an issue for some CPUs. Adaptive affects the Turbo portion of the VID stack only.

-Raja

Thanks Raj, I think I'm on the right track. The way they have it now it seems like you might be able to get away with setting it up so the overclock is stable without the extra voltage AVX/2 need added. That way the VRM will add it for you when needed as was originally intended.

Edit:
Most of my confusion with adaptive is simply down to the mysterious voltage chart Intel has devised, which is too complicated to bother publishing. You cannot simply take the voltage your board applies to auto voltage setting in BIOS and then add the extra to your known-stable manual voltage. You have to figure out how to input numbers for offsets that will cause the known-stable voltage to be applied during overclocking. There's probably more than one way to get Adaptive to work with an overclock. As in, this may not be the most efficient, but it works... I'll have to do more testing to see if it is hard capped at these voltages, but they were my desired readings:
Capture2_zps4d848de0.png~original


Btw: how do you set the SENSOR1? Is that for a different board? Looks like you could put it on the GPU and set the fans up to ramp with the GPU temp instead. Very very slick. Can't imagine being afraid of Ai Suite anymore. Shit works.
 
Last edited:
Hi,

2) Adaptive is programmed by intel. When a AVX instruction set is seen it may ramp voltage beyond the set voltage - nothing at all we can do about that.. This is known by Intel. Just how they designed the CPUs. Your other alternative is to use an offset voltage, but the issue with offset is that the change to VID affects the enitre VID stack, so that can be an issue for some CPUs. Adaptive affects the Turbo portion of the VID stack only.

-Raja

Curious about this as mentioned, on the Z87 boards you simply used Manual and used C states which allowed all the downclock/volt needed. Adaptive is a useless mode on the Z87 boards.

So Z97 boards don't behave this way anymore?
 
So Z97 boards don't behave this way anymore?

If they do, I've not yet learned the setting in the UEFI or OS that enables it, and I'm quite bummed about it. I know the proper settings for the Z87 bios by heart and there aren't really many differences to the z97 hero bios although it is quite different looking. I'm using manual mode right now but I'm considering how often I ever use anything that incorporates AVX instructions.

This could be a deal breaker for me, but the debate is between redoing everything when the board is working fine versus just running in manual mode all the time. The thing is... in manual mode--with my voltage locked down, the cpu idles at the same temp as with adaptive mode. I don't really think my system uses much energy while idle and if it stays idle long it sleeps, which is working fine for me now that I've upped my Eventual CPU input voltage. It's really bizarre that it want's to add so MUCH voltage for avx on adaptive. For giggles I tried out prime95 27.9 and it went immediately to 1.37v

**Q 4 Raja**Several days ago (Tuesday--when they did the Windows 8.1 updates, iirc), while I was still using my M6H, my system stopped waking up from deep sleep properly. It began doing a triple cold boot with a rog logo, then straight to desktop as if nothing had happened, all my apps still up, no errors in log. I tracked it down to needing to change my Eventual CPU input voltage from Auto to 1.890v. That's as low as I can go. 1.88v causes this issue. No other settings needed to be changed at all. The M7H behaves exactly the same way. It has no trouble booting to windows with Initial cpu input voltage at auto... and the other input voltage setting that is hidden away in tweakers paradise or one of those other menus near it, that one auto too. Was that the Windows update that caused that to change for me, maybe? Or is my chip "maturing"? :D any insight much appreciated as always.
 
Last edited:
Curious about this as mentioned, on the Z87 boards you simply used Manual and used C states which allowed all the downclock/volt needed. Adaptive is a useless mode on the Z87 boards.

So Z97 boards don't behave this way anymore?

Manual VCore fixes the VID the same way on Z87/Z97. Not sure why you think otherwise - some misunderstanding on your part there I think.
 
**Q 4 Raja**Several days ago (Tuesday--when they did the Windows 8.1 updates, iirc), while I was still using my M6H, my system stopped waking up from deep sleep properly. It began doing a triple cold boot with a rog logo, then straight to desktop as if nothing had happened, all my apps still up, no errors in log. I tracked it down to needing to change my Eventual CPU input voltage from Auto to 1.890v. That's as low as I can go. 1.88v causes this issue. No other settings needed to be changed at all. The M7H behaves exactly the same way. It has no trouble booting to windows with Initial cpu input voltage at auto... and the other input voltage setting that is hidden away in tweakers paradise or one of those other menus near it, that one auto too. Was that the Windows update that caused that to change for me, maybe? Or is my chip "maturing"? :D any insight much appreciated as always.

If the CPU is overclocked it will need a minimum specific delta between input VID and FIVR VID to be stable. You will need to find that yourself. I don't study Windows updates much at all so can't offer any insight there. Likely it's more to do with your OC than anything else.
 
Last edited:
Btw: how do you set the SENSOR1? Is that for a different board? Looks like you could put it on the GPU and set the fans up to ramp with the GPU temp instead. Very very slick. Can't imagine being afraid of Ai Suite anymore. Shit works.

Not sure Hero has the temp sensor header off the top of my head (gene does so it's possible) - check the manual and see if it is shown or look along the lower edge of the board for a 2-pin header marked sensor. If there is you can use it with a thermal probe cable. Then you would simply select the sensor in the drop down list called source on within the fan header's control page in Fan Xpert 3.


I will write a fan expert guide next week and post it up in the thread. SHould help a few people understand all the things FAN Xpert can do on this gen. Nothing else out there like it.
 
Last edited:
This could be a deal breaker for me, but the debate is between redoing everything when the board is working fine versus just running in manual mode all the time. The thing is... in manual mode--with my voltage locked down, the cpu idles at the same temp as with adaptive mode. I don't really think my system uses much energy while idle and if it stays idle long it sleeps, which is working fine for me now that I've upped my Eventual CPU input voltage. It's really bizarre that it want's to add so MUCH voltage for avx on adaptive. For giggles I tried out prime95 27.9 and it went immediately to 1.37v

When the CPU power gates certain sections no current is drawn (or very little) which is why you see no change in temps for the most part.

Intel designed that mechanism for a reason I suspect, possibly something to do with certain software tools needing the VID hike on the stock CPUs. It has always been up to a 0.1V boost since Z87 so nothing has changed on that front.

-Raja
 
Manual VCore fixes the VID the same way on Z87/Z97. Not sure why you think otherwise - some misunderstanding on your part there I think.

No no misunderstanding whatsoever.

Let me refer you to the way I set up Z87:
Capture_zpsc34e2f8f.jpg~original


This trick doesn't work with the new board. What's up with my Z87 trick going bybye?

Oh, it DOES have the thermal sensor header. So hmm, I need a thermal probe? Do I actually have to walk into a store and ask them where the probes are? :D
 
Last edited:
Manual VCore fixes the VID the same way on Z87/Z97. Not sure why you think otherwise - some misunderstanding on your part there I think.

Doubt that, hence my query.

Secondly, we're not talking VID here. That's what you set in BIOS and vcore is what your cores are actually using. ie: running Adaptive on a Z87 is completely useless unless you can show otherwise.
 
Doubt that, hence my query.

Secondly, we're not talking VID here. That's what you set in BIOS and vcore is what your cores are actually using. ie: running Adaptive on a Z87 is completely useless unless you can show otherwise.

For all we know that is the internal VID (have no documentation here to know what value intel reports to software or hardware Super IO). What I do know is that it needs to be a bit value to be reported to software and nothing is fast enough to show real-time voltage. Don't confuse VID for being a single value you set in UEFI - the only VID you set is the ceiling value in adaptive (sans AVX 0.1V ramp for adaptive). VID is dynamic and the value changes - otherwise how would Vcc move with power saving with your single VID, on its own? :p
 
Last edited:
No no misunderstanding whatsoever.

Let me refer you to the way I set up Z87:
Capture_zpsc34e2f8f.jpg~original


This trick doesn't work with the new board. What's up with my Z87 trick going bybye?

Oh, it DOES have the thermal sensor header. So hmm, I need a thermal probe? Do I actually have to walk into a store and ask them where the probes are? :D

1) Maybe Intel fixed what was broken on that side - who knows. As the CPU pulls so little current when idling I wouldn't worry about it either way. Proper SVID or DVID requires a VID stack and hold times to ensure the VCC has time to settle at or close to the target VID (with Voffset) before the processor goes into full load state, if this was some kind of exploit, there is a good chance it would be unstable in certain scenarios. You can try manipulating your minimum load state (1% is a bit low) and if that does not help, you might just have to live with it (or not depending upon how important you deem it to be). Officially manual VID has always been manual VID as far as I am aware, did nto even know and "exploit" existed and I'm nto the type to use exploits unless they are official, especially with something like this.

Could also be a facet of the new ME FW and microcode versions on the 9 series boards (Intel controls what happens in C-State, Offset and adaptive via the P.C.U. so this would be down to them for the most part is my guess).

2) Do what you normally do when you go to a shop and look for something you need to purchase :D
 
Last edited:
1) Maybe Intel fixed what was broken on that side - who knows. As the CPU pulls so little current when idling I wouldn't worry about it either way. Proper SVID or DVID requires a VID stack and hold times to ensure the VCC has time to settle at or close to the target VID (with Voffset) before the processor goes into full load state, if this was some kind of exploit, there is a good chance it would be unstable in certain scenarios. You can try manipulating your minimum load state (1% is a bit low) and if that does not help, you might just have to live with it (or not depending upon how important you deem it to be). Officially manual VID has always been manual VID as far as I am aware, did nto even know and "exploit" existed and I'm nto the type to use exploits unless they are official, especially with something like this.

Could also be a facet of the new ME FW and microcode versions on the 9 series boards (Intel controls what happens in C-State, Offset and adaptive via the P.C.U. so this would be down to them for the most part is my guess).

Very interesting, thanks much for your take on it. Yes--I think this is Intel's doing. I think it's fair to call it an exploit, but it is completely stable (prime95 28.1 stable at low ffts for hours), at least on the short term end (it's been working like this for me for 3+ months, I'd have to dig up the post after I first discovered it). Most people were skeptical at best until I made that screenshot.

I do think that the Intel Management Engine is involved. I don't suppose the Management Engine software from my z87 disc would work? The 1% is arbitrary, I usually just use 5%.

I just can't really see how it could be problematic, but I am no Intel Engineer, either. It is certainly a convenient solution to the question of manual vs adaptive. BOTH. I'm a little surprised you didn't know about this. I've made several posts about it but I probably should have brought it to your attention first.

If any of your guys want to test it, to me it seems like the only requirement is to put the bios volt mode to manual and then switch the windows power management profile to High Performance (balanced doesn't seem to work...) and change the minimum processor state from default 100% to 5% or whatever.

It seems like a pretty silly little trick and lots of people would notice it, but I have never seen anyone independently present this info.
 
Very interesting, thanks much for your take on it. Yes--I think this is Intel's doing. I think it's fair to call it an exploit, but it is completely stable (prime95 28.1 stable at low ffts for hours), at least on the short term end (it's been working like this for me for 3+ months, I'd have to dig up the post after I first discovered it). Most people were skeptical at best until I made that screenshot.

I do think that the Intel Management Engine is involved. I don't suppose the Management Engine software from my z87 disc would work? The 1% is arbitrary, I usually just use 5%.

I just can't really see how it could be problematic, but I am no Intel Engineer, either. It is certainly a convenient solution to the question of manual vs adaptive. BOTH. I'm a little surprised you didn't know about this. I've made several posts about it but I probably should have brought it to your attention first.

If any of your guys want to test it, to me it seems like the only requirement is to put the bios volt mode to manual and then switch the windows power management profile to High Performance (balanced doesn't seem to work...) and change the minimum processor state from default 100% to 5% or whatever.

It seems like a pretty silly little trick and lots of people would notice it, but I have never seen anyone independently present this info.

I am not referring to management engine software. I am referring to me fw and or CPU microcode.


Other than that its no big deal either way.
 
I am not referring to management engine software. I am referring to me fw and or CPU microcode.
Got it. Took me about 5 minutes to wrap my head around that. You are awfully useful, I must say--I'm slow but I get there usually.
Other than that its no big deal either way.
What I'm coming away from this conversation with is this: When we are plugged into the wall, these issues of dropping voltage are irrelevant. The power savings are just a Haswell perk which is especially useful on the mobile side. One needn't be obsessive about getting the most possible performance AND power savings from their processor because the actual savings in terms of dollars per month are essentially zero.

Because I've tested adaptive and manual modes extensively I can say with confidence that for gaming (and really anything else for that matter) manual mode is preferred because the FPS fluctuates more while the VID is dynamically changing. Using manual mode is more stable and since we have unlimited relatively inexpensive power available from our outlets it is clearly the way to go.

TL; DR: z97 manual mode 24/7 for me.
 
Hello

Adaptive mode is really not an issue IMO. Most people do not run Prime type stress tests 24/7 but instead use the system for actual programs. I have not found any programs that result in a steady increase of voltage because of AVX. Handbrake makes full use of the AVX instructions and if using a reasonable CPU voltage the increase is not something to be concerned about. Below is a screenshot of Handbrake during a fifteen minute encode. The applied voltage is nothing like one would see when running a Prime type stress test. Loaded steady state voltage is 1.287V and peak instantaneous voltage because of AVX is 1.395V


click for larger image
 
Loaded steady state voltage is 1.287V and peak instantaneous voltage because of AVX is 1.395V

1.395v is *really* high, imo, and I don't need anywhere near that kind of voltage to be stable. I don't even really use anything that contains avx instructions at all. Even using regular applications like games my voltages spike above my desired cap. As I stated above: games perform better in manual mode than adaptive, I suspect pretty much all other software will perform better with max voltage applied with no fluctuations, as well. And I'm not just saying games bench a little better. Voltage shifting in adaptive mode can cause stuttering and motion clipping in games--you can see it happening with adaptive on, and you can see it not happening with manual mode engaged. If I run handbrake, or anything else for that matter, I want to know for certain what voltage I'm at.
 
Really wish there was a simple explanation same as Z78.

Anyhow, question for Z97. Hero I believe comes with M2 PCIe connector? If I were to connect this to the PCIe3 x4, will this drop my GPU to x8 or will it remain x16? Do they share bandwidth?
 
Really wish there was a simple explanation same as Z78.

Anyhow, question for Z97. Hero I believe comes with M2 PCIe connector? If I were to connect this to the PCIe3 x4, will this drop my GPU to x8 or will it remain x16? Do they share bandwidth?

The m.2 port is separate from the PCIE ports completely, however it shares bandwidth:
M.2 Mode will make the PCIEX4_3 slot run at X2 (still useful but I can't imagine putting a soundcard here because #1 the sound out of the hero is excellent, imo and #2 I like a small case and that would put the soundcard directly above the intake to my PSU)
M.2 Mode will also disable all of the other X1 ports. I don't know what people use these for anyway but I'm probably just ignorant.

So, the m.2 port is completely SLI friendly and does not affect the 2 X16 and X8 ports.

Something you would like a lot error:
The M7H has a custom and auto fan curve editor built into the damn UEFI... it is really a lot more than I was expecting and I kinda just bumped into it a little bit ago. I don't even think it ramps the fans on boot at all anymore.

Ai Suite is pretty much irrelevant at this point and it is still causing errors for me that I cannot diagnose. I like it, but I need to make sure it's not causing any problems. I thought I had solved it but... I'm still getting a triple cold boot out of deep sleep that gives me no lead... other than dipawaymode.dll. I'm doing more testing but for now I have no idea what's up. I did some more stability testing and my overclock is rock solid so I need to make one more test of sleep mode before I take the OC off and retest to see if it is still doing it or not. If all else fails... and it is still doing it with the oc off... maybe it might be something with secure boot. I messed with secure boot and disabled it to install a unsigned driver around the time this started happening... iirc. Little hazy lol
 
Last edited:
If the CPU is overclocked it will need a minimum specific delta between input VID and FIVR VID to be stable. You will need to find that yourself. I don't study Windows updates much at all so can't offer any insight there. Likely it's more to do with your OC than anything else.

K well, Raja I think I may need further suggestion on this; I went to stock with optimized defaults and it's still happening. When waking from deep sleep my system triple boots and then goes to desktop with no error and everything saved like it should be. Strangest thing.

Not even sure where to ask for help but I'm using the z97 right so might as well be here, but it was doing it on the z87 as well. Gonna reinstall windows... surprised I'm the only one having this issue, but it's pretty much out of scope since both boards are doing it--feel free to ignore... Maybe I should go back to playing with blocks like my 4 year old.

Update: I think I may have got it. Either my APC Back-UPS Pro is faulty, or something in the windows update made it to where plugging in the usb cable that talks to it makes windows act retarded. Looking more like the latter situation.

Actually it was neither. It was the most likely scenario of all: I'm a moron. I had squeezed two oversized adapter plugs onto my ups and one of them was hanging out about 2 mm... which apparently was enough for everything to work fine... except for this bizarre behavior. I took a piece of wood and propped the UPS up so I could angle one of them down instead of up... problem solved. jeez.
 
Last edited:
I have been looking at buying the Z97 deluxe (NFC & WLC),

Some questions...
On the board's rear panel I/O, which USB ports are the chipset, and which are ASMedia's?
on the internal I/O, which usb headers are Z97's and which belong to the asmedia controller?
 
I have been looking at buying the Z97 deluxe (NFC & WLC),

Some questions...
On the board's rear panel I/O, which USB ports are the chipset, and which are ASMedia's?
on the internal I/O, which usb headers are Z97's and which belong to the asmedia controller?

1) Off the top of my head I think the two ports to left of the Ethernet hubs.

2) The FP connector ports are simply routed through an AMedia port multiplier - they are still Intel. Port multiplier works as a simple MUX/switch so no drivers required for that.

-Raja
 
Last edited:
Really wish there was a simple explanation same as Z78.
It looks like it will still work.

Seems might have to change C-state setting from Auto to enabled, and then package c-state to enabled. Wehn I put my M6H back in, it wouldn't work at first until I made this adjustment.

Edit: indeed it works. That c-states setting for package on enabled gets it working. Unfortunately in the process of doing this last motherboard swap I somehow must have damaged a trace on my naked CPU. Still works but only one memory channel works...

Edit2: pulled my chip back out and noticed some schmutz on the bottom in several different places. Apparently some clp or nt-h1 got on there when I was being sloppy. I used some thermal material remover to clean it up and set her back up fired her up and BAM both memory channels are still good. What a fucking rush. XD
 
Last edited:
Raja, if you remember my issue with the triple cold boot after waking up from sleep? I'm still chasing that issue. I've taken off my OC and run with stock settings from optimized defaults.

I think I might have a lead, but so far every time I find a fix it stops working again next day.

New thing that has caused symptom to go away. This is on the z97 Hero:
It was MCH Full Check under memory settings. It was on Auto. I switched to Enabled. Symptom disappeared for now. Will report back. I just can't imagine it's my PSU, it's nearly new and everything is rock solid stable except for this wake from sleep issue. It doesn't even seem to be deep sleep.

Also: It's pretty obvious when you have the M6H and M7H side by side that the RAM slots 'ave their colors switched. I remember from the z87 Hero manual it stated the Ram slot closer to the CPU should be populated first, and for dual channel with just 2DIMMs the other red slot on that board was proper. True to form, the M7H manual states that the red slots should be populated for 2 DIMM config, thus placing the memory conveniently a bit farther away from the CPU socket. My question: Wouldn't it have made more sense to leave the slot colors as they were and simply switch the preference so that the red slots would stand out? I guess it goes with the PCIE theme of having the red slots be for graphics cards but I have to say that it's a bit unfortunate a lot of the bling on this board ends up buried by the things we have to cram onto it. The Chipset lighting is really nice looking, but with dual GPUs it's all but invisible. The lighting should have been the Maximus VI Hero logo. Please toss my suggestion into the ring for next Hero. Personally I would even make the LED lettering white just as it is on the board right now--and have it configurable to different brightness settings and breathing option for the Halloween kids; the red lighting is often buried even further by folks' case lighting which is often red anyways. The lighting accents on the board are selling these better than the number of phases ever could, imo.
 
Last edited:
Back
Top