Z170 boot issues

I was tempted, but as bad as this issue with Gigabyte has been (assuming it is Gigabyte at fault of course), it's still nothing like as bad as the endless stream of failures I've had with Asus, so I resisted the temptation and even though they didn't really have any boards with my ideal set of features, stuck with MSI instead.

As cool as modern BIOSes are, because the frame rate is fairly low, it's not really like using a windows desktop still - it's useful to be able to use the mouse, but I'm so used the old style BIOS that I'm totally happy with that really. The only big benefit to UEFI as far as I'm concerned is much faster boot times.

samuelmorris said:

I was tempted, but as bad as this issue with Gigabyte has been (assuming it is Gigabyte at fault of course), it's still nothing like as bad as the endless stream of failures I've had with Asus, so I resisted the temptation and even though they didn't really have any boards with my ideal set of features, stuck with MSI instead.

As cool as modern BIOSes are, because the frame rate is fairly low, it's not really like using a windows desktop still - it's useful to be able to use the mouse, but I'm so used the old style BIOS that I'm totally happy with that really. The only big benefit to UEFI as far as I'm concerned is much faster boot times.

Click to expand...

True, and as you said, I've still got 10-16 months to go before I find out if my Asus turns out like yours! Hopefully not.

Hopefully not as well, in a market where there are so few decent choices these days, I'd much rather not have the opinion I do of Asus products, but every time I give them a chance within a few months I'm replacing faulty hardware again, it's too much effort nowadays. I recently tried to veer a friend off an Asus board but he stuck with it - it failed 2 days after he installed it. I try not to be the kind of person to 'I told you so' in such circumstances because it's difficult enough going through all that, but really, I'm seeing little evidence of change. That being said, statistics are now improving for Asus as their return rates are coming down, but they do still produce the least reliable motherboard on the Market (Maximus 7 Formula), so I remain very cautious.

OK well, this tale certainly took a rather interesting turn. Apologies for my late reply, a huge MS Exchange migration project at work consumed all the spare time I could muster.

I finally installed the MSI Z170A-Gaming Pro today with the original Corsair Vengeance memory and the i5 6600. Upon first boot, the PC would be stuck in a power on/power off loop, but not the 3-5 second intervals I'm used to seeing with Intel CPUs, this was half a second at most, on, off, on, off in very rapid succession. 'That's new' I thought. Playing around with cables, removing a stick of RAM again, fiddling with the front panel connector cables (as they're not labelled on the board at all with MSI, you need the manual - not happy about that!) changed nothing.
Since changing the GPU out last time did at least change the behaviour with the Gigabyte boards, I thought I'd try both using the IGP and my old HD6970, in place of the Gainward GTX970 I've been trying to use throughout.
With no graphics card installed, I was able to get the PC to boot into windows. I couldn't see anything on the screen as the board has no Displayport connector for the IGP and I was too lazy to dig out an HDMI cable. Nonetheless, the keyboard/mouse lit up to at least indicate POST, I gave it a minute, pressed the power button and around 20 seconds later the machine powered off by itself and remained off. So that clearly worked.
Curious, I reinstalled the HD6970 and was immediately met with the same issue as before with the 970 - graphics card is definitely exhonorated then.
So it's not the board, the RAM, the PSU (as verified by tests with the Gigabytes), the storage/peripherals (as all were removed for testing with the Gigabytes) or the graphics card.
What does that leave, the CPU? Even though the 6500 had issues with its Gigabyte board like the 6600, I decided to try swapping the 6500 in place of the 6600 in this board (along with using its matching CPU cooler even though they should be identical, partly just in case there was any bizarre grounding issue with one, and also because I didn't fancy mucking around with the lineup of the patterns of thermal compound that have ended up on the CPUs/coolers.
First power on, instantly the same result. OK, CPU definitely also exhonorated then.

Something I might have commented on earlier, I forget, was that the LEDs on the motherboard used to stay lit very dimly when the PC was off. I initially attributed that to the new PSU, but as it turns out I think I may have been mistaken, bear with me.
I switched the PSU off and started disconnecting everything again - first things first, the PCIe power connectors. Now bear in mind I still have the old HD6970 in the slot. I am instantly greeted with an ear-piercing beeping noise. Of course, I'd forgotten if you power up this card with a PCIe connector missing it screams at you. No matter, the PSU will discharge in a moment, I thought.
A whole minute passed. Still screaming. PSU definitely switched off at the back. I pull the cable out the back of the PSU just in case the PSU isn't switching off properly. Still screaming. After a few more seconds the noise gets higher pitched and fainter (as you would expect with capacitor discharge). However, at least 3-4 minutes later, it was still going, with mains power physically removed!
As the pitch/volume of the noise wavered, so too did the brightness of the LEDs on the board. Eventually I shut it up by pulling the graphics card out of the PCIe slot entirely, still with its displayport connector attached. That gave me an idea.

I placed the GTX970 back with the i5 6500 still in the CPU socket, switched the PSU on, reconnected the displayport connector and suddenly saw the motherboard lights come on. When I unplugged the mDP cable, they went off. I reconnected the mDP cable, powered the machine on and once again was presented with the same rapid power on/off behaviour. Out of idle curiosity I decided to unplug the displayport connector and the PC immediately resumed the rapid on/off loop. As soon as I unplugged the cable it booted into Windows immediately.
Repeated this a couple of times to be sure - PC works with mDP unplugged, doesn't work with it plugged in. Surely not?

I rummaged in the cupboard for an older, shorter DP->mDP cable that I'd put away when testing a bigger monitor (Philips BDM4065UC) that was too big for the original cable to reach and plugged that in. Motherboard LEDs remained off. Sure enough, PC booted three or four times successfully without incident, so I proceeded to replace the CPU with the 6600 again (uneventful), go back into Windows and carry on as normal.
So far, every time I have needed to reboot the PC (installing drivers etc.) I have shut the PC down fully as that used to prompt the issue with the Gigabyte boards almost immediately. So far the issue has not recurred. It's early, but it might genuinely be that the longer 3m DP->mDP cable I purchased a few months back is totally fine with a P55 board, but with Z170s it completely screws over the PCI express bus. How utterly bizarre...


tl;dr early days yet but the fault may actually have been with a mini-displayport cable

So... I guess the mDP cable was somehow shorting out and getting power to the board when it should've been off? That's... odd, to say the least. Can't say I've ever heard of such a thing. Still, if all you did was swap cables and now it works, it's pretty clear that was the source of the problem.

Yeah that's my guess, power from the monitor is coming down the mDP cable and getting through the graphics card and onto the motherboard. I have actually seen this with a PC belonging to one of our customers - a homebrew system inherited from their old IT guys where plugging an Olympus dictaphone docking station into the PC's USB as well as having its USB charger plugged into the dock at the same time caused the PC's power LED to be lit even with the machine switched off and the power button to do nothing.

On the one hand I'm embarrassed it took so long to try that, but on the other hand, how come the P55A-UD4 system is immune when none of the Z170 boards are, and how come it only trips the Gigabyte boards out on the second boot? I can only assume it's something to do with the way the PCIe controllers work in the CPU having changed over the years.

If the system's stable and works properly for the next few days I might try sticking the HD3P board in the file server as it'll just be using its IGP. I'm still nervous as I managed to get the second UD5 to fail POST once with just the IGP, but only once (in contrast with several times with the HD6970 installed), and I never repeated that test for the HD3P so fingers crossed it'll be fine... Will see what happens!

samuelmorris said:

OK well, this tale certainly took a rather interesting turn. Apologies for my late reply, a huge MS Exchange migration project at work consumed all the spare time I could muster.

I finally installed the MSI Z170A-Gaming Pro today with the original Corsair Vengeance memory and the i5 6600. Upon first boot, the PC would be stuck in a power on/power off loop, but not the 3-5 second intervals I'm used to seeing with Intel CPUs, this was half a second at most, on, off, on, off in very rapid succession. 'That's new' I thought. Playing around with cables, removing a stick of RAM again, fiddling with the front panel connector cables (as they're not labelled on the board at all with MSI, you need the manual - not happy about that!) changed nothing.
Since changing the GPU out last time did at least change the behaviour with the Gigabyte boards, I thought I'd try both using the IGP and my old HD6970, in place of the Gainward GTX970 I've been trying to use throughout.
With no graphics card installed, I was able to get the PC to boot into windows. I couldn't see anything on the screen as the board has no Displayport connector for the IGP and I was too lazy to dig out an HDMI cable. Nonetheless, the keyboard/mouse lit up to at least indicate POST, I gave it a minute, pressed the power button and around 20 seconds later the machine powered off by itself and remained off. So that clearly worked.
Curious, I reinstalled the HD6970 and was immediately met with the same issue as before with the 970 - graphics card is definitely exhonorated then.
So it's not the board, the RAM, the PSU (as verified by tests with the Gigabytes), the storage/peripherals (as all were removed for testing with the Gigabytes) or the graphics card.
What does that leave, the CPU? Even though the 6500 had issues with its Gigabyte board like the 6600, I decided to try swapping the 6500 in place of the 6600 in this board (along with using its matching CPU cooler even though they should be identical, partly just in case there was any bizarre grounding issue with one, and also because I didn't fancy mucking around with the lineup of the patterns of thermal compound that have ended up on the CPUs/coolers.
First power on, instantly the same result. OK, CPU definitely also exhonorated then.

Something I might have commented on earlier, I forget, was that the LEDs on the motherboard used to stay lit very dimly when the PC was off. I initially attributed that to the new PSU, but as it turns out I think I may have been mistaken, bear with me.
I switched the PSU off and started disconnecting everything again - first things first, the PCIe power connectors. Now bear in mind I still have the old HD6970 in the slot. I am instantly greeted with an ear-piercing beeping noise. Of course, I'd forgotten if you power up this card with a PCIe connector missing it screams at you. No matter, the PSU will discharge in a moment, I thought.
A whole minute passed. Still screaming. PSU definitely switched off at the back. I pull the cable out the back of the PSU just in case the PSU isn't switching off properly. Still screaming. After a few more seconds the noise gets higher pitched and fainter (as you would expect with capacitor discharge). However, at least 3-4 minutes later, it was still going, with mains power physically removed!
As the pitch/volume of the noise wavered, so too did the brightness of the LEDs on the board. Eventually I shut it up by pulling the graphics card out of the PCIe slot entirely, still with its displayport connector attached. That gave me an idea.

I placed the GTX970 back with the i5 6500 still in the CPU socket, switched the PSU on, reconnected the displayport connector and suddenly saw the motherboard lights come on. When I unplugged the mDP cable, they went off. I reconnected the mDP cable, powered the machine on and once again was presented with the same rapid power on/off behaviour. Out of idle curiosity I decided to unplug the displayport connector and the PC immediately resumed the rapid on/off loop. As soon as I unplugged the cable it booted into Windows immediately.
Repeated this a couple of times to be sure - PC works with mDP unplugged, doesn't work with it plugged in. Surely not?

I rummaged in the cupboard for an older, shorter DP->mDP cable that I'd put away when testing a bigger monitor (Philips BDM4065UC) that was too big for the original cable to reach and plugged that in. Motherboard LEDs remained off. Sure enough, PC booted three or four times successfully without incident, so I proceeded to replace the CPU with the 6600 again (uneventful), go back into Windows and carry on as normal.
So far, every time I have needed to reboot the PC (installing drivers etc.) I have shut the PC down fully as that used to prompt the issue with the Gigabyte boards almost immediately. So far the issue has not recurred. It's early, but it might genuinely be that the longer 3m DP->mDP cable I purchased a few months back is totally fine with a P55 board, but with Z170s it completely screws over the PCI express bus. How utterly bizarre...


tl;dr early days yet but the fault may actually have been with a mini-displayport cable

Click to expand...

Wow, what a find. Such things are a two-edged sword for me--on one side, kind of pumped to figure it out; on the other, irritated that such a simple thing caused so much grief!

Yeah. Sadly it's not even the first time such an incident has happened. Two builds before this one (Princess in the signature) I spent days changing components and trying to diagnose the fact that the PC kept powering off a few seconds after boot. With its old PSU installed (A previously used faultless Antec Earthwatts 380W unit), it made a horrific squealing noise in time with the HDD LED - every time the light flash, it'd emit this noise. The shop where I bought the MB/CPU/RAM/GPU from for the upgrade told me to replace the PSU which shut the noise up but the power fault remained, the system would rarely boot and when it did it shut down abruptly a few seconds after. Changing motherboards didn't help (I still have the other spare one on my shelf, 3 years later, never used it or sold it, whoops!) but I eventually tracked it down to one of the SATA cables. Replaced that and instantly the issue disappeared with no other changes. I'm guessing the cable was shorting the SATA data to ground so the PSU was simply reacting to a short that was to the frequency of the activity on the SATA interface, shutting down entirely when it got too much. That was a week's diagnosis I could have done without!

After this month (and yes it has taken that long!) I'm really growing to suspect cables a lot more when things go wrong... Guessing I owe Gigabyte an apology, but then that first board was definitely DOA with memory problems - none of the other boards suffered that particular fault. I should also thank MSI for having a board that made the issue easier to diagnose with its flashing LEDs going nuts when the cable was inserted. That's the one I'm keeping, because I've proven it works, and also because the lighting effect is quite nice - even if I loathe the fact that there's a huge red LED inside the ethernet socket that makes it almost impossible to see the link status LEDs either side!

I am having issues myself with the GA-Z170N-Gaming 5 Mini ITX
The issues are extremely inconsistent and weird.

Latest BIOS (F3)

i5 6600 skylake
16GB DDR4 2133 ADATA
MSI Gaming GTX 970

Basically, most of the time the machine will boot fine. However, if Windows crashes or something, any time the machine goes down unexpectedly, I'll get stuck in a power-on/power-off loop (3-5 seconds). Waiting a few minutes or clearing the CMOS seems to fix it.

Secondly, and this I find really strange, I can make changes in the BIOS to simple things (like boot order, fan speeds, etc). However, the *moment* I touch a CPU setting, (for instance, I want to disable Turbo Boost) and I save and exit, I'm back to boot loops again, and the board eventually recovers itself with the backup bios and goes back to default settings. I cannot set any power states or anything. All CPU settings must be at default or I get boot loops. Even changing memory settings does not cause an issue.

I've tried disabling fast boot, enabling "power loading", and various settings, and those settings will change fine, but if I touch ANY CPU setting, I'm sitting here back at square one with boot loops again.

I've tried switching up the RAM sticks in different slots, one stick at a time, etc and nothing there seems to help.

The odd thing is, once I get it up and running with defaults, it seems stable in Windows for the most part (past couple of days I've had it crash while gaming but a Windows issue seemed to be at fault there and it is now solved).

I have no idea what to think. Gigabyte could not offer much help. The machine was a Syber, purchased from Cyberpower (which I regret, I should have just built it myself, the Syber cases are heat traps and the machine is only stable with the top and bottom of the case completely off).

Temps are fine too in current config, so I don't think that's the main issue. Seems to be the Gigabyte board, but it's so odd that it *can* be stable for days at a time, and then unstable when something changes, or when I try to modify any BIOS setting. *sigh*

If the machine came from cyberpower then it's on them to fix that. Again, perhaps it's something I'm too easy to jump to but in that situation I would try a board swap. I haven't owned one myself but everything I've heard about cyberpower machines is negative, usually things like suspect low quality cases and PSUs etc. That being said, unless it is a PSU problem (which sounds a little unlikely) then I can't see what they could have done wrong there to cause that...

xx0xx said:

I am having issues myself with the GA-Z170N-Gaming 5 Mini ITX
The issues are extremely inconsistent and weird.

...

Basically, most of the time the machine will boot fine. However, if Windows crashes or something, any time the machine goes down unexpectedly, I'll get stuck in a power-on/power-off loop (3-5 seconds). Waiting a few minutes or clearing the CMOS seems to fix it.

Secondly, and this I find really strange, I can make changes in the BIOS to simple things (like boot order, fan speeds, etc). However, the *moment* I touch a CPU setting, (for instance, I want to disable Turbo Boost) and I save and exit, I'm back to boot loops again, and the board eventually recovers itself with the backup bios and goes back to default settings. I cannot set any power states or anything. All CPU settings must be at default or I get boot loops. Even changing memory settings does not cause an issue.

I've tried disabling fast boot, enabling "power loading", and various settings, and those settings will change fine, but if I touch ANY CPU setting, I'm sitting here back at square one with boot loops again.

I've tried switching up the RAM sticks in different slots, one stick at a time, etc and nothing there seems to help.

The odd thing is, once I get it up and running with defaults, it seems stable in Windows for the most part (past couple of days I've had it crash while gaming but a Windows issue seemed to be at fault there and it is now solved).

I have no idea what to think. Gigabyte could not offer much help. The machine was a Syber, purchased from Cyberpower (which I regret, I should have just built it myself, the Syber cases are heat traps and the machine is only stable with the top and bottom of the case completely off).

Temps are fine too in current config, so I don't think that's the main issue. Seems to be the Gigabyte board, but it's so odd that it *can* be stable for days at a time, and then unstable when something changes, or when I try to modify any BIOS setting. *sigh*

Click to expand...

I joined the forum specifically to comment on this.
I have a similar - but not identical - Gigabyte Z170 motherboard .

Mine is a GA-Z170XP-SLI . Whenever I disable "Turbo Boost" in the BIOS, it immediately goes into a boot loop, where the power actually shuts down, then it comes on for 4 seconds, then shuts down, etc. Nothing is ever displayed on the screen.
The only way out is to short the CMOS reset pins.

I have tried multiple versions of the BIOS - F3 that came with my board, and F5 (latest). The issue is the same.

FYI, here is my config :
Intel i5-6600k CPU
NH-D14 cooler (with 2011 kit, separately purchased)
32 GB DDR4-3000 (2x16GB) . Kit is Patriot PV432G300C6K .
Gigabyte GTX750 Ti . Model GV-N75TOC-2GI .

I have tried removing the GPU also, and using only the iGPU, and the problem still happens when disabling Turbo Boost . The only other thing I could do is to remove one stick of RAM, and try all 4 RAM slots. I have not done so, as I doubt this would help. I have tested both memory sticks with Memtest, FYI, and they passed a 12 hour test which was several passes, even when running at their rated DDR4-3000 speed. So I know both RAM sticks are good.

The boot loop problem is easily reproducible even hitting F7 - "Load Optimized Defaults" in the BIOS, and only disabling the Turbo Boost .

I also have 3 more PCI-E and PCI cards, but I have taken them out and the problem does happen even without them also. But I will list them just in case :

- Syba SY-PEX30016 1394a/1394b PCI-E 2.0 x1 card (Need this for my professional audio interface)
- Syba SI-PEX40063 SATA/eSATA PCI-E 2.0 x2 card . Need this to support external port multiplier enclosure, as Intel SATA controller won't. Also, I filled all 6 SATA ports on the motherboard already.
- VIA USB 2.0/1394/IDE PCI card. I am only using this for the USB 2.0 ports, as the Gigabyte mobo doesn't have enough 2.0 ports on the backplate, though it has plenty of 3.0/3.1. I didn't want to fill the 3.x ports with my USB 1.x/2.0 devices. And yes I have 2.0 hubs, 5 hubs in fact, but I didn't want to chain all those 2.0 hubs to a single 2.0 motherboard port. One mobo 2.0 port has to be for the keyboard to get into the BIOS. I believe keyboard on hub won't work.

FYI, I have seen several other problems with the board besides the Turbo Boost disable boot loop :

1) with a Symbios 53C1010 Ultra160 SCSI PCI controller. The card's SCSI ROM BIOS takes about 2 minutes to execute, whereas it only took about 20 seconds on my previous motherboard. And the ROM only runs the very first boot after the card is inserted - and never again on subsequent boots, regardless of BIOS settings about storage option ROMs. This could be an ASMedia 1083 PCI bridge issue, or again a Gigabyte BIOS issue.

2) when using both internal GPUs and external GPUs in combination. I need to enable the internal GPU in order to have QuickSync enabled for video encoding in Windows. I have 3 displays, two of which require dual-link DVI and are on the discrete 750Ti GPU, and the third display is on the internal GPU. My old 750Ti can only encode H.264 in hardware, but not H.265. But QuickSync on the Skylake 6600k with the Intel Graphics 530 is capable of doing H.265 hardware encoding. Now, if only Youtube would allow H.265 clip uploads ...

The problem is that I have seen frequent lockups related to the internal GPU. After Windows boots, the third display, which is on the internal GPU, gets filled with a single random solid color , and the whole machine hangs at that point. Not even the RESET button on the case works ! The only way out is to hold the power button for 5 seconds to turn the power off, and restart. And I have seen this happen at stock clock and RAM settings, without any OC at all. At least when it happens, it does so early in the boot process, not in the middle of doing actual work with the machine. If I disable the internal GPU and only use the discrete GPU, and run all 3 displays off it, then I have no problem with the occasional lockups on boot. I will eventually upgrade my 750Ti GPU to one that is H.265 capable, and make the Quicksync moot, and at that point this will be less of an issue as I will be able to disable the internal GPU.

I think the BIOS on our Gigabyte Z170 motherboards is extremely buggy. I only used Gigabyte motherboards for the last 8 years in many builds, at least 10 builds for myself and family members, and was generally very happy, but I'm extremely disappointed with this Z170XP-SLI model I got, and it seems that you are seeing very similar issues. I have built my own PCs for 23 years, and I'm a programmer day for about the same amount of time, so I'm not exactly a newbie to this.

I'm at the point where I'm considering switching for another motherboard of a different brand, if I could be sure that the 3 above problems with disabling Turbo Boost in the BIOS, the SCSI PCI controller ROM support, and the internal GPU lockups would not occur.

That said, once I learned what Gigabyte BIOS setting traps to avoid, I will say that I have achieved an OC of 4.5 GHz on my 6600k with my Gigabyte board that is prime95 28.7 stable in a 21 hour blend test.
This involves setting the vcore to 1.39V in the BIOS, and setting the Turbo Boost ratios to 45 for all 4 options (1 core, 2 cores, 3 cores, 4 cores) since I can't disable turbo boost.

The CPU VID in HWmonitor does not show this value - it is much lower, like in the 1.23-1.25V range, and is dynamic. But somehow, setting the 1.39V vcore value in the BIOS (or in Easytune) is still required to achieve the stable OC - all lower values caused the Prime95 test to fail with errors. I really wish I could disable the Turbo Boost so the VID was actually static. Even disabling EIST and all C-states cannot force the VID to be static.

As an aside, with the OC I achieved, my RAM is running as the rated 3000 MHz with its XMP profile - I did not try to run it any faster. In the BIOS, the RAM voltage shows 1.35V in one place, and 1.2V in another screen, sigh. Looks like another Gigabyte bug.

tl;dr Gigabyte has BIOS bugs, disabling Turbo Boost causes boot loops. This gets in the way of achieving a good OC. And there are several other Gigabyte Z170 issues with PCI and iGPU.

Wow yeah sounds like an identical issue. CyberPower can fix it / accept a return at this point but there will be horrendous shipping costs (restocking fees if I don't want it just fixed) and other things that will get in the way.

I'm just marking it as one of the worst buying decisions I've made in some time. All of the other components seem rock solid. I am skeptical of the PSU, though. I think I'm going to disassemble the machine and put it in another case with maybe even another PSU and see what happens. Lastly, I will probably just replace the motherboard myself.

Not sure what to do at this point. All of my options suck.

I also fully admit that I'm dumb for not returning it in the first 30 days. But... the problem is within the first couple of days I got the machine stable (temps were the issue originally) and so I felt I had a stable running machine I could keep running for a long time. One of my recent game crashes (due to a PAGEFAULT error) sent it back into boot-loops again, so I'm at square one. I really think replacing the Mobo and staying away from Gigabyte will be the best option.

The is the first issue with Gigabyte I've experienced, but when I look up "boot loop issue" and "Gigabyte", it is unbelievable how many results out there that there are for these boards with a similar issue

Hi,

xx0xx said:

Wow yeah sounds like an identical issue. CyberPower can fix it / accept a return at this point but there will be horrendous shipping costs (restocking fees if I don't want it just fixed) and other things that will get in the way.

I'm just marking it as one of the worst buying decisions I've made in some time. All of the other components seem rock solid. I am skeptical of the PSU, though. I think I'm going to disassemble the machine and put it in another case with maybe even another PSU and see what happens. Lastly, I will probably just replace the motherboard myself.

Not sure what to do at this point. All of my options suck.

I also fully admit that I'm dumb for not returning it in the first 30 days. But... the problem is within the first couple of days I got the machine stable (temps were the issue originally) and so I felt I had a stable running machine I could keep running for a long time. One of my recent game crashes (due to a PAGEFAULT error) sent it back into boot-loops again, so I'm at square one. I really think replacing the Mobo and staying away from Gigabyte will be the best option.

The is the first issue with Gigabyte I've experienced, but when I look up "boot loop issue" and "Gigabyte", it is unbelievable how many results out there that there are for these boards with a similar issue

Click to expand...

I'm still within the first 30 days - bought my Gigabyte mobo on Black friday. It's a pain to rebuild everything with a new mobo.

Fry's has an Asus Z170-AR board which is about $35 more than what I paid - and only one legacy PCI slot vs 2 on the Gigabyte.

But when I google "Asus Z170 boot loop", I see plenty of matches too. I wonder if the issue is with the Z170 chipset and not just a Gigabyte BIOS issue.

xx0xx said:

I am having issues myself with the GA-Z170N-Gaming 5 Mini ITX
The issues are extremely inconsistent and weird.

Latest BIOS (F3)

i5 6600 skylake
16GB DDR4 2133 ADATA
MSI Gaming GTX 970

Basically, most of the time the machine will boot fine. However, if Windows crashes or something, any time the machine goes down unexpectedly, I'll get stuck in a power-on/power-off loop (3-5 seconds). Waiting a few minutes or clearing the CMOS seems to fix it.

Secondly, and this I find really strange, I can make changes in the BIOS to simple things (like boot order, fan speeds, etc). However, the *moment* I touch a CPU setting, (for instance, I want to disable Turbo Boost) and I save and exit, I'm back to boot loops again, and the board eventually recovers itself with the backup bios and goes back to default settings. I cannot set any power states or anything. All CPU settings must be at default or I get boot loops. Even changing memory settings does not cause an issue.

I've tried disabling fast boot, enabling "power loading", and various settings, and those settings will change fine, but if I touch ANY CPU setting, I'm sitting here back at square one with boot loops again.

I've tried switching up the RAM sticks in different slots, one stick at a time, etc and nothing there seems to help.

The odd thing is, once I get it up and running with defaults, it seems stable in Windows for the most part (past couple of days I've had it crash while gaming but a Windows issue seemed to be at fault there and it is now solved).

I have no idea what to think. Gigabyte could not offer much help. The machine was a Syber, purchased from Cyberpower (which I regret, I should have just built it myself, the Syber cases are heat traps and the machine is only stable with the top and bottom of the case completely off).

Temps are fine too in current config, so I don't think that's the main issue. Seems to be the Gigabyte board, but it's so odd that it *can* be stable for days at a time, and then unstable when something changes, or when I try to modify any BIOS setting. *sigh*

Click to expand...

Hello, I think this is a known issue for Gigabyte and they already got the fix for this problem no system boot when disabled turbo boost, please ask Gigabyte for the latest beta bios for this fix.
For Z170N-Gaming 5 they have bios F4b for the fixed.

pchealth 2 said:

Hello, I think this is a known issue for Gigabyte and they already got the fix for this problem no system boot when disabled turbo boost, please ask Gigabyte for the latest beta bios for this fix.
For Z170N-Gaming 5 they have bios F4b for the fixed.

Click to expand...

I'm running the latest BIOS F5 for my GA-Z170XP-SLI, which was released on October 13 :
http://www.gigabyte.com/products/product-page.aspx?pid=5496#bios

I just opened a support ticket with Gigabyte for this issue on my board.

It looks like BIOS F4beta for your Z170N-Gaming 5 is not public yet.

Re: the core voltage, it seems Hwmonitor 1.28 does not show the core voltage properly under CPU VID.

However, CPU-Z 1.74 shows core voltage properly. The value shown there matches the Vcore set in the BIOS, as well as Vcore set in Easytune.

madbrain said:

I'm running the latest BIOS F5 for my GA-Z170XP-SLI, which was released on October 13 :
http://www.gigabyte.com/products/product-page.aspx?pid=5496#bios

I just opened a support ticket with Gigabyte for this issue.

Re: the core voltage, it seems Hwmonitor 1.28 does not show the core voltage properly under CPU VID.

However, CPU-Z 1.74 shows core voltage properly.

Click to expand...

Hello, what is the bios version you are using, I think they already got the beta bios for this model I think. You can avail for this beta bios for checking.
By the way you say CPU Z 1.74, Easytune and MB Vcore read correctly, only HWmonitor reads incorrect.
Maybe you may check if there any new version of HWmonitors.
And also, as far as I know, Gigabyte has bundle a software from Intel XTU on their DVD driver disk.
You can verified the actual Vcore under this software. You can check it on the Intel XTU lower left corner, you will see a graphical flow chart and under this column you can see a small screw plier for setting.
Under setting you may check the Vcore to show on the graph.
http://www.gigabyte.com/products/product-page.aspx?pid=5496#utility

pchealth 2 said:

Hello, what is the bios version you are using, I think they already got the beta bios for this model I think. You can avail for this beta bios for checking.

Click to expand...

Where could I find the beta BIOS for my board ? It is not posted on the Gigabyte web site support pages.

By the way you say CPU Z 1.74, Easytune and MB Vcore read correctly, only HWmonitor reads incorrect.
Maybe you may check if there any new version of HWmonitors.
And also, as far as I know, Gigabyte has bundle a software from Intel XTU on their DVD driver disk.
You can verified the Vcore under this software.
http://www.gigabyte.com/products/product-page.aspx?pid=5496#utility

Click to expand...

I have checked, and HWMonitor 1.28 is the latest version.

I have the Gigabyte edition of Intel XTU, and it is up to date according to the update check, but it is showing yet another vcore value, that does not match what's in the BIOS/CPU-Z/Easytune.

In fact, XTU is showing all kinds of clearly bogus values - like 3408 Watts max Turbo Boost power.

You can avail the latest beta bios on their Gigabyte Technical support link.
Or you may check the latest beta bios on this link for Gigabyte:
http://forums.tweaktown.com/gigabyte/28441-gigabyte-latest-beta-bios.html
I have here the latest beta bios F4B on my Z170N-Gaing 5 and have correct reading on Intel XTU, CPU Z and Bios, not try HW monitors.
I will try to download HWmonitor for checking.
After checking HWmonitor 1.28, it seem to be different from CPU Z 1.74.
It seems HWmonitor does not have corrected like CPU Z, as far as I know CPU Z has correction as before.

pchealth 2 said:

You can avail the latest beta bios on their Gigabyte Technical support link.
Or you may check the latest beta bios on this link for Gigabyte:
http://forums.tweaktown.com/gigabyte/28441-gigabyte-latest-beta-bios.html

Click to expand...

Thanks, but I couldn't find any beta BIOS there for the Z170XP-SLI .

I have here the latest beta bios F4B on my Z170N-Gaing 5 and have correct reading on Intel XTU, CPU Z and Bios, not try HW monitors.
I will try to download HWmonitor for checking.

Click to expand...

FYI, here are some screenshots on my system with current 4.5 GHz OC and 1.39 Vcore set in BIOS.

First EasyTune (correct) :



Next, CPU-Z 1.74 (also correct) :


Next, Hwmonitor (incorrect) :


Finally, XTU (incorrect, many bogus numbers) :


In addition, there is a review of my GA-Z170XP-SLI with an OC guide that claims the reviewer was able to disable Turbo Boost .

See :
https://forums.overclockers.co.uk/showthread.php?t=18690668



The reviewer was using BIOS F3 . I tried reverting to that version, and still get the boot loop when disabling Turbo Boost .

The reviewer was using a 6700k also, whereas I have a 6600k . Maybe this BIOS bug is only seen with the 6600k . Or maybe I have a bad board.

I also do not see a beta bios for my board, the GA-Z170N-Gaming 5

Only these:

GA-Z170MX-Gaming 5 - F4b
GA-Z170N-WIFI - F7
GA-Z170X-Gaming 5 - F5d

Contacting Gigabyte again and seeing if they can provide a BIOS. I'm letting them know that others are seeing similar issues with Gaming series boards.

Hello Madbrain, I think you should submit a ticket to Gigabyte and get the latest beta bios for Z170XP-SLI. The latest bios should have fixed this disabled Turbo boost.Please do mentioned about HWMoinitor also as CPU Z has corrected the reading on GIgabyte before and may have not change the setting for HW monitor.

I RMA'd my Z170 Gigabyte 7 mobo bc of boot issues. Asus might have terrible warranty service but if you get a working board then you're good bc their bios is top notch

pchealth 2 said:

Hello Madbrain, I think you should submit a ticket to Gigabyte and get the latest beta bios for Z170XP-SLI. The latest bios should have fixed this disabled Turbo boost.Please do mentioned about HWMoinitor also as CPU Z has corrected the reading on GIgabyte before and may have not change the setting for HW monitor.

Click to expand...

I opened a ticket on Wednesday, but have yet to receive a reply from them. The ticket is still in the support queue, being processed.

It looks like HWMonitor was last updated in July, whereas CPU-Z was updated much more recently. I hope the fix will make it in HWMonitor soon.

Gigabyte sent me the F4B for the Z170N-Gaming 5 which is the model I have, apparently this one isn't public yet. I will update on how it goes.

I am here to report good news! So far with the F4B BIOS they sent, it is working wonderfully.

- Turbo boost and power options can now be changed without causing boot loops
- There are extra options now available before (mouse speed in BIOS, scrollwheels now work in BIOS, etc)

I am very excited about this new development. We'll see how stable it is over time. I had some separately unstable RAM that I took out, I'm going to put it back in and try it with the new BIOS and see how that goes. If it's still an issue it may just be the one stick of RAM.

But definitely recommend F4B for GA-Z170N Gaming 5. Like night and day.

I asked them to make it publicly downloadable if possible. If they don't, I'll see if I can get it to someone to post on Tweaktown's list

Gigabyte sent me the F6D BIOS for my GA-Z170XP-SLI .

I used the Gigabyte AtBIOS to flash it from Windows.

And ... Lo and behold, the entire machine shut down in the middle of the flash process, and immediately went into a loop where it powers up for 4 seconds, then powers off, and so on.

I powered down the system (turned off the PSU switch in the back) and then turned the machine back on. Now, there is no longer a boot loop. The fans come on. But, nothing else. No video, no keyboard LED activity.

My guess is that their flash program does not work properly when the CPU is OC.

So much for the "Dual BIOS" feature that Gigabyte advertises for situations where the BIOS flashing process is interrupted. It's supposed to automatically recover and restore the factory version.

http://www.gigabyte.us/microsite/55/tech_081226_dualbios.htm

DualBIOS is listed as a feature on my Z170XP-SLI :
http://www.gigabyte.com/products/product-page.aspx?pid=5496#sp

I guess a completely dead motherboard is a good reason to return it to Fry's. It's 7:30pm and I Fry's closes at 9pm, so I still have time to take the box apart, pack the board, and get another Z170 motherboard. I don't think it will be another Gigabyte, however.

never update bios in windows, always do it from either build in bios update option or dos.

Windows bios update is really bad bad idea.

madbrain said:

Gigabyte sent me the F6D BIOS for my GA-Z170XP-SLI .

I used the Gigabyte AtBIOS to flash it from Windows.

And ... Lo and behold, the entire machine shut down in the middle of the flash process, and immediately went into a loop where it powers up for 4 seconds, then powers off, and so on.

I powered down the system (turned off the PSU switch in the back) and then turned the machine back on. Now, there is no longer a boot loop. The fans come on. But, nothing else. No video, no keyboard LED activity.

My guess is that their flash program does not work properly when the CPU is OC.

So much for the "Dual BIOS" feature that Gigabyte advertises for situations where the BIOS flashing process is interrupted. It's supposed to automatically recover and restore the factory version.

http://www.gigabyte.us/microsite/55/tech_081226_dualbios.htm

DualBIOS is listed as a feature on my Z170XP-SLI :
http://www.gigabyte.com/products/product-page.aspx?pid=5496#sp

I guess a completely dead motherboard is a good reason to return it to Fry's. It's 7:30pm and I Fry's closes at 9pm, so I still have time to take the box apart, pack the board, and get another Z170 motherboard. I don't think it will be another Gigabyte, however.

Click to expand...

Don't give up hope quite yet. Clear the CMOS using the pins and wait a while. I've had this exact same thing happen with this board, and eventually I was able to coax it alive and it overwrote the bios from the backup. It takes a bit of back and forth / CMOS clearing / removing power to get it, but you may still be able to get it to restore the BIOS.

Also seconding not updating BIOS from Windows, it's always seemed to be more unstable that way. QFlash has worked flawless for me on Gigabyte boards that have it. If yours doesn't, use the other non-Windows process.

nekrosoft13 said:

never update bios in windows, always do it from either build in bios update option or dos.

Windows bios update is really bad bad idea.

Click to expand...

I had actually had gone back and forth between BIOS versions on this mobo, from F3 to F5, and back, several times, always flashing from Windows. I didn't have any issues. I'm not sure if Gigabyte sent me a bad BIOS file or if the flashing went bad due to the OC.

It was certainly very strange to see the machine power off in the middle of the flashing - not just reboot. I would suspect a mismatch BIOS first. But it still shouldn't have caused a power off. It wasn't a power surge - my other PC stayed on, as did everything else in the house.

xx0xx said:

Don't give up hope quite yet. Clear the CMOS using the pins and wait a while. I've had this exact same thing happen with this board, and eventually I was able to coax it alive and it overwrote the bios from the backup. It takes a bit of back and forth / CMOS clearing / removing power to get it, but you may still be able to get it to restore the BIOS.

Click to expand...

How long exactly was "a while". I shorted the pins for over 30s, both without the PSU connected to the wall (per instructions) and with.

I also took out every add-on card and left just the USB keyboard and one monitor connected to the iGPU. I had no luck reviving it.

Also seconding not updating BIOS from Windows, it's always seemed to be more unstable that way. QFlash has worked flawless for me on Gigabyte boards that have it. If yours doesn't, use the other non-Windows process.

Click to expand...

Well, for what it's worth, I have been updating about a dozen Gigabyte motherboards from Windows for myself and family for at least the last decade, never had a problem, but I only one of those mobos was ever overclocked . I especially like that their program always find the properly matching BIOS version for the particular motherboard revision, as they have many. Check out all the revisions of the GA-990FXA-UD3 ...

I think it is likely the case that Gigabyte support actually sent a mismatched BIOS file for my mobo.

In any case, I gave up on the Z170XP-SLI mobo. I wasn't happy with it overall for several other reasons, not just the Turbo Boost BIOS problem, which I listed in my first post in the forum earlier in this thread.

I only had until December 29 to return it, but I will be out of the country from the 25th on, so really only had a few days left to decide to keep the board or not.

The board was returned to Fry's as defective shortly before the Fry's closing time. The store will send it back to Gigabyte. I now own an Asus Z170A-AR . This was a bit more expensive motherboard (about $30 more) with a very similar feature set. I hope its BIOS is better than the Gigabyte motherboard . And I will be able to determine if the other issues with the PCI SCSI controller, and the lockups with the iGPU, are issues with Z170 chipset issues and ASMedia PCI bridge in general, or were rather Gigabyte issues on that particular Z170XP-SLI motherboard.

Now on to dinner and a fun late night piecing it back together.

madbrain said:

I had actually had gone back and forth between BIOS versions on this mobo, from F3 to F5, and back, several times, always flashing from Windows. I didn't have any issues. I'm not sure if Gigabyte sent me a bad BIOS file or if the flashing went bad due to the OC.

It was certainly very strange to see the machine power off in the middle of the flashing - not just reboot. I would suspect a mismatch BIOS first. But it still shouldn't have caused a power off. It wasn't a power surge - my other PC stayed on, as did everything else in the house.

Click to expand...

doesn't matter if you didn't, it's always more prone for issues. you just got lucky so far.

any hardware instability, windows software, unstable OC, bsod, driver issues, other software can interfere, taking extra unnecessary risks.

nekrosoft13 said:

doesn't matter if you didn't, it's always more prone for issues. you just got lucky so far.

any hardware instability, windows software, unstable OC, bsod, driver issues, other software can interfere, taking extra unnecessary risks.

Click to expand...

It seems the process for BIOS updating on the Z170 is much longer and different than on any other boards I owned.

Even on my new mobo, the Asus Z170-AR, when I flashed the latest BIOS version from the UEFI BIOS, it was an affair where the machine powered itself off and back on about 5 times, to update various firmware components separately like ME Firmware. It did succeed in the end. This is clearly not your grandfather's BIOS update process, however.

Maybe I didn't give the Gigabyte board a fair share of tries to recover. Perhaps it would eventually have done so. I was concerned about the wear and tear on all my disks and other devices powering themselves on and off so many times in a short period of time. Especially as I have seen so many boot loops already with the Gigabyte board.

Anyway, I'm still glad I don't have that Gigabyte motherboard anymore in the end.
I haven't seen any such boot loop at all with the new Asus board.

Pros of the new Asus Z170-AR board over the Gigabyte Z170XP-SLI :
- it has a DisplayPort for 4K @ 60 Hz. The Z170XP-SLI did not. It means I may not have to buy a new GPU if I replace a single monitor with a 4K model. My GTX750Ti doesn't have a DisplayPort, so no 4K at 60 Hz is possible.
- it has an SPDIF optical out, which was sorely missing on the Gigabyte motherboard
- has one more standard fan header, as well as waterpump EXT_FAN headers. Now all but one of my 2 CPU fans + 5 case fans are hooked up to the motherboard fan controller.
- I haven't seen any hard lockup after booting Windows related to using the iGPU, like I did with the Gigabyte board. This was one of the 3 main problems I was having with the Gigabyte. So, good news - I think this was a Gigabyte issue, or a defective board, and not a Z170 chipset issue. This comforts me that I did the right thing in replacing my Gigabyte motherboard with the Asus .

Cons of the Z170-AR board :
- some of the Asus software conflicts with my Firewire PCI-E card & external audio interface. I don't know which program it is yet. Uninstalling all Asus software in Windows solved the problem.
- another software issue - Asus AI Charger Plus was crashing on startup . I don't need this anyway.
- selecting XMP profile to get my RAM to run at its rated DDR4-3000 results in no soft reboot - the machine just hangs with all the fans at 100%. Only a cold boot with the power button allows it to boot, and then everything is fine and CPU-Z reports the RAM correctly at the right speed in Windows. The XMP profile was not doing this on the Gigabyte motherboard. I will reach out to both Asus and Patriot to figure out if there is a manual BIOS setting I can use to solve this. I will also try a lower RAM speed. There is no issue with soft reboot at the stock speed of DDR4-2133 .
- there is only one legacy PCI slot vs 2 on the Z170XP-SLI . I have 2 PCI cards I would like to install - one PCI SCSI controller, and one multi-function board to get more USB 2.0 ports on the back panel. For now I'm running with the later. But when I switch to the PCI SCSI card, I can just plug in some more cheap USB hubs to solve the shortage of USB 2.0 ports issue.

Other notes :
- as far as I can tell, everything is fully stable at stock settings with this board, which is more than I could say for the Gigabyte board, so this was definitely a step forward. No boot loops, no lockups shortly after Windows starts.
- there was no need to rerun the Microsoft genuine software in Windows 7 when switching to a different brand of Z170 motherboard. Everything else was kept the same (CPU, GPU, and 3 other add-on cards).
- the RAID volume I had created on the Gigabyte board with my two old 128GB SSDs still worked fine with the Asus board . First boot was a bit messy with all the device changes, but after about 10 minutes Windows figured out everything, then I went on to uninstall all the Gigabyte software and install the Asus software. All the drivers for the iGPU, storage, USB 3.1 and NICs were the same, so need to install any new ones. Actually I already had newer versions installed than what was on the Asus mobo DVD.
- I have not tried the PCI SCSI controller yet, but will do so soon.
- I have not tried to disable Turbo Boost yet - the Asus BIOS layout is nice but very new to me, so I don't know where that option is yet.
- I have not tried to OC the CPU yet . I was hoping to OC the RAM first (see above) and then the CPU, but perhaps I need to do both at once.
- the Asus board has an old-fashioned jumper to allow running the CPU at something other than stock voltage. I was surprised to see that. I set the jumper to enable, but I'm still running the CPU at stock voltage and clock in the BIOS.

madbrain said:

- I have not tried to disable Turbo Boost yet - the Asus BIOS layout is nice but very new to me, so I don't know where that option is yet.

Click to expand...

So, I found the setting in the Z170-AR BIOS, and it works just fine - no boot loop.

However, doing this doesn't appear to be the way to go for OC . The CPU clock remained at the base 3.5 GHz regardless of the other clock settings I had set for 4.5 GHz.
So, it looks like I actually need to keep Turbo on to OC the CPU.
Nevertheless, I'm glad there isn't a "BIOS trap" that causes a boot loop, like there was on the Gigabyte board.

FYI, once I disabled Speedstep, Turbo and all the C-states, HWMonitor 1.28 stopped showing the CPU VID altogether . But it shows CPU Vcore under the motherboard. The max value appears to correspond to what I set in the BIOS, 1.39V. The min value is 0.688V though, which doesn't make much sense.

I will have to read the threads about the Asus Z170-A, which is nearly identical to the Z170-AR to see how others have done their OC as far as BIOS settings go. Probably later today.

I am quite glad my boot loops are gone. Those are incredibly frustrating, especially when after clearing CMOS it was still wanting to loop some. Now I've not been able to get it to loop anymore. I've tried all sorts of settings and so forth and I think it's finally stable.

*phew* Crisis Averted.

xx0xx said:

I am quite glad my boot loops are gone. Those are incredibly frustrating, especially when after clearing CMOS it was still wanting to loop some. Now I've not been able to get it to loop anymore. I've tried all sorts of settings and so forth and I think it's finally stable.

*phew* Crisis Averted.

Click to expand...

Happy for you ! I wish the BIOS that Gigabyte support sent me had worked as well for me as the one you got. There were so many times I had to clear CMOS with my Z170XP-SLI . This sounds like a case where Gigabyte went to market too early, with an unfinished product.

I have yet to reset CMOS even once with the new Z170-AR . But I haven't played with every single BIOS setting yet. The automatic OC tool from Asus gave me a pretty lame CPU OC of 4.1 GHz for all 4 cores, but at least it worked. The amazing part is that it can run this way with 4 out of the 5 case fans at 0rpm (and their respective blue LEDs completely off!) and the CPU temp isn't going above 53C during a long video encode. I think this is because I turned on the EPU. The machine is amazingly silent this way though, sounds more like a laptop than a desktop. I think this is because I turned on the "EPU" option in the AI Suite. Turns out this is related to energy efficiency - I don't really want that.

The advanced mode in the 5-way optimization failed - it turns on XMP and that results in no soft boot which causes the AI Tweaker to fail even in the very first OC test. There may be a way to make it run without doing a RAM OC, though.

madbrain said:

Happy for you ! I wish the BIOS that Gigabyte support sent me had worked as well for me as the one you got. There were so many times I had to clear CMOS with my Z170XP-SLI . This sounds like a case where Gigabyte went to market too early, with an unfinished product.

I have yet to reset CMOS even once with the new Z170-AR . But I haven't played with every single BIOS setting yet. The automatic OC tool from Asus gave me a pretty lame CPU OC of 4.1 GHz for all 4 cores, but at least it worked. The amazing part is that it can run this way with 4 out of the 5 case fans at 0rpm (and their respective blue LEDs completely off!) and the CPU temp isn't going above 53C during a long video encode. I think this is because I turned on the EPU. The machine is amazingly silent this way though, sounds more like a laptop than a desktop. I think this is because I turned on the "EPU" option in the AI Suite. Turns out this is related to energy efficiency - I don't really want that.

The advanced mode in the 5-way optimization failed - it turns on XMP and that results in no soft boot which causes the AI Tweaker to fail even in the very first OC test. There may be a way to make it run without doing a RAM OC, though.

Click to expand...

For those who care, the XMP issue was due to me not RTFM - my 2 DIMMs were in slots A1 and B1 rather than A2 and B2 where they should have been according to the manual. When they were moved, XMP worked just fine with no soft reboot problem.

However, I have discovered one annoying issue - the Firewire audio dropouts I was seeing earlier have been traced not just to Asus Windows software, but to any kind of OC setting done in the BIOS, whether RAM or CPU, even just enabling XMP. It's bad enough that Windows can't play the startup sound without a scratching pop . With the fastest OC configs I have tried, the sound actually gets cut off. I pretty much can't play any audio off my Firewire interface for more than a few seconds. If I revert to stock, everything is fine, I can run Prime95, a long video encode, and HD/SSD benchmarks like HDTach, all concurrently, and still not have my audio playback be disrupted by a single pop, let alone cut off completely. It took me a lot of attempts at settings changes and reboots to trace this. I definitely did not expect that, and did not see that issue with the Gigabyte motherboard. The Firewire interface has to be 100% reliable in order for me to be able to record audio. So for now, it looks like I have to run everything at stock, both CPU and RAM I will definitely have to reach out to Asus about this one.

The long boot with the SCSI controller ROM OTOH appears to be related to the ASMedia PCI bridge, on the other hand, it's not either an Asus or Gigabyte issue.

I found a spare USB 2.0 bracket to attach to one of the mobo headers, so no need for the second PCI card to get those ports. And my case (CM HAF-XM) conveniently has one extra PCI bracket slot so it doesn't take away an expansion slot - since I filled all but 2 of the slots - and I also have an RS232 COM port bracket installed for my X10.

madbrain said:

For those who care, the XMP issue was due to me not RTFM - my 2 DIMMs were in slots A1 and B1 rather than A2 and B2 where they should have been according to the manual. When they were moved, XMP worked just fine with no soft reboot problem.

However, I have discovered one annoying issue - the Firewire audio dropouts I was seeing earlier have been traced not just to Asus Windows software, but to any kind of OC setting done in the BIOS, whether RAM or CPU, even just enabling XMP. It's bad enough that Windows can't play the startup sound without a scratching pop . With the fastest OC configs I have tried, the sound actually gets cut off. I pretty much can't play any audio off my Firewire interface for more than a few seconds. If I revert to stock, everything is fine, I can run Prime95, a long video encode, and HD/SSD benchmarks like HDTach, all concurrently, and still not have my audio playback be disrupted by a single pop, let alone cut off completely. It took me a lot of attempts at settings changes and reboots to trace this. I definitely did not expect that, and did not see that issue with the Gigabyte motherboard. The Firewire interface has to be 100% reliable in order for me to be able to record audio. So for now, it looks like I have to run everything at stock, both CPU and RAM I will definitely have to reach out to Asus about this one.

The long boot with the SCSI controller ROM OTOH appears to be related to the ASMedia PCI bridge, on the other hand, it's not either an Asus or Gigabyte issue.

I found a spare USB 2.0 bracket to attach to one of the mobo headers, so no need for the second PCI card to get those ports. And my case (CM HAF-XM) conveniently has one extra PCI bracket slot so it doesn't take away an expansion slot - since I filled all but 2 of the slots - and I also have an RS232 COM port bracket installed for my X10.

Click to expand...

Good luck to you with your new Asus Z170A-R, I seem to find difficulty on it bios setting so I don't prefer Asus.

pchealth 2 said:

Good luck to you with your new Asus Z170A-R, I seem to find difficulty on it bios setting so I don't prefer Asus.

Click to expand...

Thanks. I had much less difficulty with the Asus BIOS than the Gigabyte BIOS, especially when it came to boot loops.

I was able to achieve the same 4.5 GHz CPU & cache OC on the Asus that I had on the Gigabyte board.

I narrowed down the issue with the Firewire to a problem with the RAM OC, not CPU OC. The problem occurs whenever the RAM is set to a speed greater than 2533 MHz. I posted in an Asus forum about this at https://vip.asus.com/forum/view.asp...anguage=en-us&page=1&board_id=1&model=Z170-AR . So far, no helpful replies. I tried 2 different Firewire cards - one PCI and one PCI-E, and same issues. Tried 2 different sets of RAM sticks also, including one that's on the Asus QWL - no difference. The same Firewire card also runs in an X99 machine at DDR4-2800 without issue ... So this Firewire issue with high RAM speed seems to be specific to Z170/Skylake ...

Anyway I have been running Prime95 on the Z170-AR with CPU/cache at 4.5 GHz and RAM at 2533 MHz, and all seems to be fine with Vcore at 1.39V - I set the same Vcore I had on the Gigabyte motherboard. I will give it 24 hours - 10 hours so far.
My attempt at 4.6 GHz failed so far 1h35 into the Prime95 28.7 blend test. I'm not really sure if it's worth all that much more time trying to get to 4.6 . If I could get 4.8 or 5.0, yes. No one has succeeded at 5.0 yet though, so that's a long shot.

Over at another forum, there is a table of statistics . Seems a lot of people are achieving 4.6 GHz on the 6600k, a few 4.8 GHz, and one 4.7 and another 4.9 . But only a few are using P95 v28.7 to test.

I may try a different way to OC the RAM later on using a different multiplier and BCLK, to see if I can achieve > 2533 MHz RAM speed without losing the Firewire. It sure would be nice to at least achieve the 3000 MHz that my sticks are rated for.

There is one user that achieved 4.8 GHz in P95 and may be worth trying to replicate at 4.8 GHz on my system, once I am convinced my 4.5 GHz OC is stable.

mccalas 6600k 100 48 4.8 3.5 1.425 1.416 1 H100i GTX P95 v28.7 7hr L521B469 2666 16-18-18-36 @ 1.2v Gigabyte z170 Gaming 7 LLC High Yes

My Patriot PV432G300C6K RAM is rated for 3000 MHz 16-16-16-36 at 1.35V. Currently I'm running at 2533 16-16-16-36 at 1.35V. I probably can run it at 1.2V only that speed, and maybe with lower latency timings too - just not sure which ones.

Perhaps by lowering the RAM voltage, as well as lowering cache OC, and adding a little bit of CPU vcore I could successfully push it to 4.8 GHz in P95. My temps at 4.5 GHz don't seem to be a problem, but they didn't seem so either at 4.6 GHz, yet P95 still managed to report a rounding error at 4.6 .

Notting informed me that the F4C bios from Gigabyte for the Z170N-Gaming 5 is now available from Gigabyte directly.
Since I'm stable with F4B, I probably will not flash any further, but this probably contains the same fixes as F4B plus other tweaks.

I wish they would be more detailed in their release notes.