[SOLVED] No Post after Overclock Ryzen 7 2700X - Asus Rog Strix X470-i Gaming

[captainchaos79]

Desperately need help, because I'm in it up to my neck! I built this PC a few months back and it's been running like a dream. I had overclocked the CPU with great success and without any problems. I had updated the bios recently with Asus EZ bios update and wanted to take advantage of some of the new features in the bios for overclocking. I managed to overclock to 4.35mhz, tested it in Cinebench everything was stable and achieved my highest score. Rebooted tested again in Cinebench but this time with a crash, rebooted to Bios, upped the voltage slightly and retested in Cinebench with success. I wanted to see if I could slightly lower the voltage in the bios but I think here In my haste put in a higher voltage. Tested in Cinebench it crashed, then after reboot my PC won't post. I have done many overclocks and if it crashes a reboot usually sorts it and at worst a Cmos reset. Everything appears to be powering up, fans are on, there's power to the mobo, Ram and GPU, everything is saying it's working but without it posting. I have taken the ram out and tried 1 stick at a time, I have removed the GPU and reseated it, I tried the onboard VGA, still no post. I have even unplugged everything from the board and when I mean everything I mean everything including the CPU, reset cmos and plugged it all back in, still the same result no post. The Q LED sequence on the Mobo is Orange, Red(CPU), Yellow(Ram), then flickers White(VGA), then repeats the sequence like it's trying to post, then the White LED stays on permanently and according to the manual the corresponding LED remains lit until the problem is solved therefore supposed to be a VGA problem. I don't believe it to be the GPU after a CPU overclock, plus it won't post even without the GPU in and using the onboard Graphics. I hope someone can help, because I'm out of ideas.

Mobo - Asus Rog Strix X470-i Gaming

CPU - Ryzen 7 2700X

GPU - EVGA GTX 1080i Hybrid SC2

Ram - G.Skill Trident Z RGB 32gb 3200MHz

 

[Dunlop0078]

Reset the CMOS. This will reset the BIOS to default and should allow you to post and redo your settings. The process is in your motherboards manual, on that board I believe it just involves moving a jumper to a different set of pins then back.

 

[captainchaos79]

Reset the CMOS. This will reset the BIOS to default and should allow you to post and redo your settings. The process is in your motherboards manual, on that board I believe it just involves moving a jumper to a different set of pins then back.

Thanks for getting back to me.
Believe me when I say I have tried resting the Cmos in as many ways possible, including removing the Mobo battery. On this Mobo it's 2 pins that you just bridge with something metal. 99.9% of the time with previous overclocks that have crashed the cmos reset works and you can get back to Bios, but not this time, that includes the removal of the battery which I never had to do before.

 

[alextheblue]

I tried the onboard VGA, still no post.
…
plus it won't post even without the GPU in and using the onboard Graphics.

Tip: There are NO onboard graphics. NONE. The onboard graphics on an AM4 system are only functional if you have an APU installed with integrated graphics. For example a 2200G. Your 2700X (and indeed all Ryzen branded chips currently on the market without a "G") do NOT have ANY graphics. The connectors on the mainboard you've tested are nonfunctional without an APU installed. They're only present so that someone can use the same motherboard for APUs and have a connection for their monitor. For you, they're decoration.

Now then, why do you think a failed GPU is impossible? Graphics cards do fail randomly. I'm not saying that is the problem but don't dismiss it off hand. Gaming GPUs can and do go kaput seemingly at random. If you have an older working GPU that would be ideal, or if you have access to another PC you can test your GPU with that might help. For that matter I wouldn't even eliminate the PSU as a possibility, and I'd double-check all your GPU-PSU connections too while you're poking around.

With that being said I don't know what your cooling setup is like, but I can tell you from experience that I no longer run fans at very low speeds at idle while overclocking, even if my CPU temp is good. I use tower coolers and I find they don't adequately cool the board unless you're moving a certain minimum amount of air, and overclocking/overvolting exacerbates this problem. So aiming for ultra-silent was causing the occasional stability problem. In my case tweaking the fan settings in my BIOS made my system boot 100% consistently (previously it would occasionally hiccup before Windows loaded and AI suite took over), and I also made sure that AI Suite settings (while less aggressive than what I set in the BIOS) run the fans at a reasonable level at idle, and it ratchet up a bit more aggressively than previously under a load. I know you're already beyond this point and can't tinker with fan speeds, but it's something to keep in mind for the future. Hopefully no harm has been done to the board, but if there's an issue you can probably RMA the board.

I know how frustrating it can be when something goes wrong. There's a lot of crazy things that can go wrong. Good luck, I hope it's just a bad GPU.

 

[captainchaos79]

I tried the onboard VGA, still no post.
…
plus it won't post even without the GPU in and using the onboard Graphics.

Tip: There are NO onboard graphics. NONE. The onboard graphics on an AM4 system are only functional if you have an APU installed with integrated graphics. For example a 2200G. Your 2700X (and indeed all Ryzen branded chips currently on the market without a "G") do NOT have ANY graphics. The connectors on the mainboard you've tested are nonfunctional without an APU installed. They're only present so that someone can use the same motherboard for APUs and have a connection for their monitor. For you, they're decoration.

Now then, why do you think a failed GPU is impossible? Graphics cards do fail randomly. I'm not saying that is the problem but don't dismiss it off hand. Gaming GPUs can and do go kaput seemingly at random. If you have an older working GPU that would be ideal, or if you have access to another PC you can test your GPU with that might help. For that matter I wouldn't even eliminate the PSU as a possibility, and I'd double-check all your GPU-PSU connections too while you're poking around.

With that being said I don't know what your cooling setup is like, but I can tell you from experience that I no longer run fans at very low speeds at idle while overclocking, even if my CPU temp is good. I use tower coolers and I find they don't adequately cool the board unless you're moving a certain minimum amount of air, and overclocking/overvolting exacerbates this problem. So aiming for ultra-silent was causing the occasional stability problem. In my case tweaking the fan settings in my BIOS made my system boot 100% consistently (previously it would occasionally hiccup before Windows loaded and AI suite took over), and I also made sure that AI Suite settings (while less aggressive than what I set in the BIOS) run the fans at a reasonable level at idle, and it ratchet up a bit more aggressively than previously under a load. I know you're already beyond this point and can't tinker with fan speeds, but it's something to keep in mind for the future. Hopefully no harm has been done to the board, but if there's an issue you can probably RMA the board.

I know how frustrating it can be when something goes wrong. There's a lot of crazy things that can go wrong. Good luck, I hope it's just a bad GPU.

Thanks for the advice, I appreciate it.
I had no idea my CPU didn't have built in graphics and that's what effects the onboard Graphics.
I know where your coming from with the GPU but surely after a CPU overclock crash it would be unlikely. If the board was bricked can and would power still be getting to all other components. Also could it just be that the board has fried the PCie slot. I really hope it's not the GPU it was expensive, at least if it's the Mobo I can RMA it.
Damn it, coz I don't have another card to test or another PC I'm gonna have to take it somewhere to get it tested.
Thanks for the tips on the cooling as well. Although I am using an NZXT H200i Mini ITX case, because of the heat generated by the CPU and GPU, I have the NZXT water cooling for the CPU and the GPU is already water cooled and my CPU temps were always pretty good, even when overclocking they would be no higher than 68c trest tested and normal use around 40c, the GPU again usually around 40c and benchmarked around 50c

 

[alextheblue]

Thanks for the advice, I appreciate it.
I had no idea my CPU didn't have built in graphics and that's what effects the onboard Graphics.
I know where your coming from with the GPU but surely after a CPU overclock crash it would be unlikely. If the board was bricked can and would power still be getting to all other components. Also could it just be that the board has fried the PCie slot. I really hope it's not the GPU it was expensive, at least if it's the Mobo I can RMA it.
Damn it, coz I don't have another card to test or another PC I'm gonna have to take it somewhere to get it tested.
Thanks for the tips on the cooling as well. Although I am using an NZXT H200i Mini ITX case, because of the heat generated by the CPU and GPU, I have the NZXT water cooling for the CPU and the GPU is already water cooled and my CPU temps were always pretty good, even when overclocking they would be no higher than 68c trest tested and normal use around 40c, the GPU again usually around 40c and benchmarked around 50c

The GPU could have failed coincidentally. I was just saying if the board was indicating it was hanging up on VGA, you need to consider the possibility. The GPU isn't under warranty? CPUs rarely fail unless you really jacked the voltage up to an insane level or physically damaged it. I would tend to say the mainboard is the most likely, but I don't know what you ran the CPU at prior to the no-POST.

The thing about water cooling is that it isn't great at cooling mainboard components. If you're water cooling the GPU and the CPU in a small case, how much airflow are other components getting? In particular the VRMs. If all you monitor is GPU/CPU die temps, you could have issues without even knowing it. Just as an example a stock OEM downdraft HSF cools VRMs better than a typical high-end tower cooler, without additional aid. Water cooling can be awesome (especially for dense builds) but you need to make sure you've got good overall airflow in your chassis - especially when pushing components.

I'm not saying that is the case but you need to be aware of it. A 4.35Ghz all-core 24/7 overclock is pretty aggressive for a daily driver 2700X. At recommended voltage levels (sub 1.4) TH only got their sample stable at 4.2, I mean obviously samples vary and some people push harder but just something to keep in mind. Do you know what voltage you were running it at?

 

[captainchaos79]

Thanks for the advice, I appreciate it.
I had no idea my CPU didn't have built in graphics and that's what effects the onboard Graphics.
I know where your coming from with the GPU but surely after a CPU overclock crash it would be unlikely. If the board was bricked can and would power still be getting to all other components. Also could it just be that the board has fried the PCie slot. I really hope it's not the GPU it was expensive, at least if it's the Mobo I can RMA it.
Damn it, coz I don't have another card to test or another PC I'm gonna have to take it somewhere to get it tested.
Thanks for the tips on the cooling as well. Although I am using an NZXT H200i Mini ITX case, because of the heat generated by the CPU and GPU, I have the NZXT water cooling for the CPU and the GPU is already water cooled and my CPU temps were always pretty good, even when overclocking they would be no higher than 68c trest tested and normal use around 40c, the GPU again usually around 40c and benchmarked around 50c

The GPU could have failed coincidentally. I was just saying if the board was indicating it was hanging up on VGA, you need to consider the possibility. The GPU isn't under warranty? CPUs rarely fail unless you really jacked the voltage up to an insane level or physically damaged it. I would tend to say the mainboard is the most likely, but I don't know what you ran the CPU at prior to the no-POST.

The thing about water cooling is that it isn't great at cooling mainboard components. If you're water cooling the GPU and the CPU in a small case, how much airflow are other components getting? In particular the VRMs. If all you monitor is GPU/CPU die temps, you could have issues without even knowing it. Just as an example a stock OEM downdraft HSF cools VRMs better than a typical high-end tower cooler, without additional aid. Water cooling can be awesome (especially for dense builds) but you need to make sure you've got good overall airflow in your chassis - especially when pushing components.

I'm not saying that is the case but you need to be aware of it. A 4.35Ghz all-core 24/7 overclock is pretty aggressive for a daily driver 2700X. At recommended voltage levels (sub 1.4) TH only got their sample stable at 4.2, I mean obviously samples vary and some people push harder but just something to keep in mind. Do you know what voltage you were running it at?

With some of the new options in the Asus Bios updates, there are a few new tweaks which includes one for water cooled CPU that gives you a bit more scope. I had the CPU ratio at 42 and then the BCLK at 101, I was surprised myself that I got such a high overclock but it was stable at around 1.40V The increase in voltage helped, the crazy thing is I meant to drop it back down a touch to see if it would stay stable but that's were I think I put a higher voltage instead of a lower. But I take on board what you have said about the VRM and previously about having the fans on the faster settings rather than silent settings when overclocking. I do also have a 120mm fan that blows from the top of the case but whether that's enough air flow in a smaller case I'm not sure. Well my current situation maybe a prime example of not enough air flow. The blower style water cooled card I have makes a difference, I know the 1080Ti's do get hot but most of the heat is blown out of the rear of the case which makes a huge difference. Anyway none of this solves my current problem at least for now but your right on considering the GPU if the facts are the VGA LED coming on the Mobo, and it's always a process of elimination in these situations, so I'm of to my local PC repair tomorrow to get the Mobo and GPU tested. Just out of interest is there any way of monitoring the VRM temps?

 

[alextheblue]

Just out of interest is there any way of monitoring the VRM temps?

Digital temp probe or thermal imaging. Or just make sure a fan is blowing on them and keep under 1.4V.

 

[captainchaos79]

Just to update this and finalise this thread for anyone who may have a similar problem!
I did indeed either cook my CPU or got unlucky and the CPU just died. Either way I was overclocking so likely it was my own fault. So the L.E.D on my Mobo hung on white to indicate a problem with VGA which is kind of right because without a working CPU there is no VGA and therefore no POST. Much to my relief it wasn't my very expensive GPU (although I thought it would unlikely to be that that had failed) and the CPU was still under warranty so RMA'd it. Worked straight away with the replacement CPU. The lesson here is don't play around with voltage when overclocking and if you do increase voltage then only a very small increase at the most, unless your experienced of course, and as alextheblue pointed out cooling is paramount even if you have liquid cooling like myself, other areas like VRM will need cooling as well especially in smaller iTX cases like mine. So make sure you ramp up the fans and water cooling fans to max when overclocking.

 

[alextheblue]

Thank you for the followup. Yeah I suspect the voltage was accidentally bumped a tad too high. Every CPU architecture has its quirks and limits. Like how you have to be mindful of RAM voltage on many recent Intel chips. With the Zen chips (at least the current models) 1.4V seems to be the safe limit, and most articles I've seen recommend staying under that for any long-term builds.

The lesson here is don't play around with voltage when overclocking and if you do increase voltage then only a very small increase at the most, unless your experienced of course, and as alextheblue pointed out cooling is paramount even if you have liquid cooling like myself, other areas like RMA* will need cooling as well especially in smaller iTX cases like mine.

*VRMs (typo)

I'm glad it wasn't anything major, and that you were able to RMA it.