Question Ryzen 7 5700x won't stop shutting PC off when gaming.

[rcald2000]

Minipiwi. That's so interesting because I didn't think your issue was related to temperature at all. The CPU temperature range that you quoted didn't seem high enough to even throttle, let alone reboot the system. And your system wasn't rebooting, but was shutting off. As for the VRM overheating, that would seem unlikely with a Ryzen 7 5800X (105 watt TDP), let alone a R7 5700X (65 watt TDP). I'm aware that your board has no heatsinks on the VRM, but you're not even pushing an all -core workload, simply gaming. Would you be willing to take a photo of the inside of your entire build, showing the layout and airflow, uploading it to imgur and sharing the link here?

My theory was that high transient wattage spikes of your graphics card were triggering protection on your power supply, resulting in a shutdown. I've read of many PSUs doing this with RTX 3080 (and above) GPUs , and your RTX 3070 Ti has the same nominal wattage as that card. I recall that you're running a Corsair RMx 1000, and if memory serves me, they provide only 8 pin cable PCIe 6+2 pin, dual connectors. As an extra precaution, I would run one 8 pin cable from the PSU to graphics card, and leave the dual connector unattended. Then I would run a second 8 pin cable from the PSU to the graphics card, and then leave that dual connector unattended. If I'm wrong, then I'm wrong. Regardless, I still recommend that you open a ticket. If you believe that the issue is motherboard VRM related then open a ticket with Asrock. And if PSU related then open a ticket with Corsair.

If your issue truly is related to high temperatures, then I'll thankful that you've zeroed in on the cause with the help of the community. Be cool.

* notes *

google query: rtx 3080 corsair shutdown

RTX 3080 causing PC shutdown while gaming. No restart, hard shutdown.
htt-ps://www.reddit.com/r/buildapc/comments/wx3u14/rtx_3080_causing_pc_shutdown_while_gaming_no/

Rtx 3080 Shutdown Issues
htt-ps://www.techpowerup.com/forums/threads/rtx-3080-shutdown-issues.291811/

RTX 3070 Ti getting shutdowns with Seasonic Focus GX-750 (Mfg. date Oct 2020) (Also with Corsair RM1000x [2021])
*htt-ps://linustechtips.com/topic/1374580-rtx-3070-ti-getting-shutdowns-with-seasonic-focus-gx-750-mfg-date-oct-2020-also-with-corsair-rm1000x-2021/page/2/

https://www.asrock.com/mb/AMD/B450M-HDV R4.0/index.asp#BIOS

 

[Minipiwi]

Minipiwi. That's so interesting because I didn't think your issue was related to temperature at all. The CPU temperature range that you quoted didn't seem high enough to even throttle, let alone reboot the system. And your system wasn't rebooting, but was shutting off. As for the VRM overheating, that would seem unlikely with a Ryzen 7 5800X (105 watt TDP), let alone a R7 5700X (65 watt TDP). I'm aware that your board has no heatsinks on the VRM, but you're not even pushing an all -core workload, simply gaming. Would you be willing to take a photo of the inside of your entire build, showing the layout and airflow, uploading it to imgur and sharing the link here?

My theory was that high transient wattage spikes of your graphics card were triggering protection on your power supply, resulting in a shutdown. I've read of many PSUs doing this with RTX 3080 (and above) GPUs , and your RTX 3070 Ti has the same nominal wattage as that card. I recall that you're running a Corsair RMx 1000, and if memory serves me, they provide only 8 pin cable PCIe 6+2 pin, dual connectors. As an extra precaution, I would run one 8 pin cable from the PSU to graphics card, and leave the dual connector unattended. Then I would run a second 8 pin cable from the PSU to the graphics card, and then leave that dual connector unattended. If I'm wrong, then I'm wrong. Regardless, I still recommend that you open a ticket. If you believe that the issue is motherboard VRM related then open a ticket with Asrock. And if PSU related then open a ticket with Corsair.

If your issue truly is related to high temperatures, then I'll thankful that you've zeroed in on the cause with the help of the community. Be cool.

* notes *

google query: rtx 3080 corsair shutdown

RTX 3080 causing PC shutdown while gaming. No restart, hard shutdown.
htt-ps://www.reddit.com/r/buildapc/comments/wx3u14/rtx_3080_causing_pc_shutdown_while_gaming_no/

Rtx 3080 Shutdown Issues
htt-ps://www.techpowerup.com/forums/threads/rtx-3080-shutdown-issues.291811/

RTX 3070 Ti getting shutdowns with Seasonic Focus GX-750 (Mfg. date Oct 2020) (Also with Corsair RM1000x [2021])
*htt-ps://linustechtips.com/topic/1374580-rtx-3070-ti-getting-shutdowns-with-seasonic-focus-gx-750-mfg-date-oct-2020-also-with-corsair-rm1000x-2021/page/2/

https://www.asrock.com/mb/AMD/B450M-HDV R4.0/index.asp#BIOS

Yeah for the picture here's the link, sorry for bad quality

http://imgur.com/a/5WIkwEJ

View: https://imgur.com/gallery/5WIkwEJ

As for what you said about the wattage, I'm not sure, I honestly don't even know what a VRM is so all of this is new but if it turns out it's not temps then I will try this. Like I've said it no longer happens since taking the side panel off and putting my massive fan on it.

 

[rcald2000]

That case looks very spacious and airy. I can't see the front fans in the photo, but it really surprises me that there is a heat issue. With that being said, is there space outside the backside of the case? Or it it against the wall? But hopefully your issue is closed to be fully resolved.

 

[Karadjgne]

And your system wasn't rebooting, but was shutting off.

Same thing, depending on whether Op has the bios set to restart after power-loss or not. A shutdown and restart, or just shutdown.

I recall that you're running a Corsair RMx 1000,

Then it's not transient spikes, especially from a 3070ti. Corsair RMx use a doubled up OCP, it's got 2 trigger points depending on the situation. They run a lower limit/long duration trip and a higher limit/short duration trip. A transient spike generally lasts less than 10ms, so doesn't affect the lower limit OCP which is looking for the standard high current draws, it affects the higher limit trip that's looking for extreme high, short duration spike loads. Corsair was the only company to do so, other companies, such as Seasonic, have redesigned their psus to match.

A 3070ti isn't going to trip OCP on an RMx 1000w, and even if by some fluke it did, it would require physically turning off the psu or unplugging it and you'd hear an audible click as the OCP resets. Absolutely no reboot or autostart possible without that power-off reset.

VRM is the Voltage Regulatory Mosfets, those are the little square/round cap looking things on the motherboard to the left of the socket. With low end motherboards they often don't have an included heatsink, but don't generally require one as the expected power use will be far below the rating of the mosfets. Such as the use of a 65w class Ryzen cpu. If you intended on using a high-core, 105w class Ryzen like a 5900x or 5950x, then a better grade of board, with VRM heatsinking would be advised, especially if trying to run with PBO enabled.

First thing I would try is reinstall Windows. This can be done without any personal data loss or loss of files. It's not uncommon for cpu swaps to mess with windows settings, some of which remain set at the previous cpu levels. It's like being used to roller skates, and one day you switch to ice skates, your eyes see ice, you brain knows absolutely you are on ice with ice skates, but for some odd reason your feet behave like you are still on roller skates, and you end up on your butt real quick and in a hurry.

 

[Minipiwi]

That case looks very spacious and airy. I can't see the front fans in the photo, but it really surprises me that there is a heat issue. With that being said, is there space outside the backside of the case? Or it it against the wall? But hopefully your issue is closed to be fully resolved.

I have no fans in the front what so ever

 

[Minipiwi]

Same thing, depending on whether Op has the bios set to restart after power-loss or not. A shutdown and restart, or just shutdown.

Then it's not transient spikes, especially from a 3070ti. Corsair RMx use a doubled up OCP, it's got 2 trigger points depending on the situation. They run a lower limit/long duration trip and a higher limit/short duration trip. A transient spike generally lasts less than 10ms, so doesn't affect the lower limit OCP which is looking for the standard high current draws, it affects the higher limit trip that's looking for extreme high, short duration spike loads. Corsair was the only company to do so, other companies, such as Seasonic, have redesigned their psus to match.

A 3070ti isn't going to trip OCP on an RMx 1000w, and even if by some fluke it did, it would require physically turning off the psu or unplugging it and you'd hear an audible click as the OCP resets. Absolutely no reboot or autostart possible without that power-off reset.

VRM is the Voltage Regulatory Mosfets, those are the little square/round cap looking things on the motherboard to the left of the socket. With low end motherboards they often don't have an included heatsink, but don't generally require one as the expected power use will be far below the rating of the mosfets. Such as the use of a 65w class Ryzen cpu. If you intended on using a high-core, 105w class Ryzen like a 5900x or 5950x, then a better grade of board, with VRM heatsinking would be advised, especially if trying to run with PBO enabled.

First thing I would try is reinstall Windows. This can be done without any personal data loss or loss of files. It's not uncommon for cpu swaps to mess with windows settings, some of which remain set at the previous cpu levels. It's like being used to roller skates, and one day you switch to ice skates, your eyes see ice, you brain knows absolutely you are on ice with ice skates, but for some odd reason your feet behave like you are still on roller skates, and you end up on your butt real quick and in a hurry.

Ahhh I see, sorry for the basic question but how would I reinstall windows? And I do believe it to be temps. Was playing cod rn for an hour or two and I only had one crash, that being a CTD cause the game crashed but no shutdown as of yet. So I firmly do believe it has to do with temperatures. I only have to fans, one on the back and one on top, none in the front and the panel to the right has no space.

 

[Minipiwi]

That case looks very spacious and airy. I can't see the front fans in the photo, but it really surprises me that there is a heat issue. With that being said, is there space outside the backside of the case? Or it it against the wall? But hopefully your issue is closed to be fully resolved.

Sorry I didn't add this but yes there is, I'd say close to a foot of space.

 

[Minipiwi]

Same thing, depending on whether Op has the bios set to restart after power-loss or not. A shutdown and restart, or just shutdown.

Then it's not transient spikes, especially from a 3070ti. Corsair RMx use a doubled up OCP, it's got 2 trigger points depending on the situation. They run a lower limit/long duration trip and a higher limit/short duration trip. A transient spike generally lasts less than 10ms, so doesn't affect the lower limit OCP which is looking for the standard high current draws, it affects the higher limit trip that's looking for extreme high, short duration spike loads. Corsair was the only company to do so, other companies, such as Seasonic, have redesigned their psus to match.

A 3070ti isn't going to trip OCP on an RMx 1000w, and even if by some fluke it did, it would require physically turning off the psu or unplugging it and you'd hear an audible click as the OCP resets. Absolutely no reboot or autostart possible without that power-off reset.

VRM is the Voltage Regulatory Mosfets, those are the little square/round cap looking things on the motherboard to the left of the socket. With low end motherboards they often don't have an included heatsink, but don't generally require one as the expected power use will be far below the rating of the mosfets. Such as the use of a 65w class Ryzen cpu. If you intended on using a high-core, 105w class Ryzen like a 5900x or 5950x, then a better grade of board, with VRM heatsinking would be advised, especially if trying to run with PBO enabled.

First thing I would try is reinstall Windows. This can be done without any personal data loss or loss of files. It's not uncommon for cpu swaps to mess with windows settings, some of which remain set at the previous cpu levels. It's like being used to roller skates, and one day you switch to ice skates, your eyes see ice, you brain knows absolutely you are on ice with ice skates, but for some odd reason your feet behave like you are still on roller skates, and you end up on your butt real quick and in a hurry.

Sorry that this is an old post, I'm still having this issue two weeks later and I no longer think it's temps. I made a video since when it shuts down without warning I never get to see the temps so I proped my phone and record. What I got was that through my gaming session it was at 61 for CPU 45 for gpu. A mere second before the total shutdown showed me 63 degress for CPU and 47 for gpu. I have no idea where to go from here.

 

[drivinfast247]

Yeah for the picture here's the link, sorry for bad quality

http://imgur.com/a/5WIkwEJ

View: https://imgur.com/gallery/5WIkwEJ

As for what you said about the wattage, I'm not sure, I honestly don't even know what a VRM is so all of this is new but if it turns out it's not temps then I will try this. Like I've said it no longer happens since taking the side panel off and putting my massive fan on it.

So you have ZERO issues when the side panel is off and a fan blowing on board??

 

[rcald2000]

My theory was that high transient wattage spikes of your graphics card were triggering protection on your power supply, resulting in a shutdown. I've read of many PSUs doing this with RTX 3080 (and above) GPUs , and your RTX 3070 Ti has the same nominal wattage as that card. I recall that you're running a Corsair RMx 1000, and if memory serves me, they provide only 8 pin cable PCIe 6+2 pin, dual connectors. As an extra precaution, I would run one 8 pin cable from the PSU to graphics card, and leave the dual connector unattended. Then I would run a second 8 pin cable from the PSU to the graphics card, and then leave that dual connector unattended. If I'm wrong, then I'm wrong. Regardless, I still recommend that you open a ticket. If you believe that the issue is motherboard VRM related then open a ticket with Asrock. And if PSU related then open a ticket with Corsair.

Have you tried any of my suggestions above? Also, in the photo I could swear that only one 8 pin PCIe cable is connected to your RTX 3070 Ti. Would you mind taking and posting another photo of the inside of your build, but with the room lights on? Speaking of your GPU, what is the precise model of Zotac RTX 3070 Ti? Please post a link to the specific product page. Also, just out of curiosity, did you end up upgrading the CPU cooler? I don't personally think your issue is related to CPU temps, but the information might be helpful.

 

[Minipiwi]

Have you tried any of my suggestions above? Also, in the photo I could swear that only one 8 pin PCIe cable is connected to your RTX 3070 Ti. Would you mind taking and posting another photo of the inside of your build, but with the room lights on? Speaking of your GPU, what is the precise model of Zotac RTX 3070 Ti? Please post a link to the specific product page. Also, just out of curiosity, did you end up upgrading the CPU cooler? I don't personally think your issue is related to CPU temps, but the information might be helpful.

I have not no, I will try to reinstall windows but wouldn't I have to buy windows ? Like I said I got this oc prebuilt so all that stuff already came in it and I have no idea how to even install windows. Here is a link to the card I bought, I got it on April 4th from Amazon: https://www.amazon.com/gp/aw/d/B09CML48LD?psc=1&ref=ppx_pop_mob_b_asin_title
And here is the links for the pictures, hope they're better:

http://imgur.com/a/qqsmVNX

View: https://imgur.com/gallery/qqsmVNX

And I did not upgrade me cooler yet no, I've had doubts that it would actually fix the problem especially after what I saw last night. That was with the panel taken off and the fan on it too at max speed.

 

[rcald2000]

Understood. Try my suggestions with the cable. One cable to one PCIe socket on the GPU. Then a second cable to the other PCIe socket on the GPU; two entirely separate cables. Let me know how it goes. Good job on the photos, by the way. And if after that, the problem reoccurs then please open up a ticket with Corsair.

 

[Minipiwi]

Understood. Try my suggestions with the cable. One cable to one PCIe socket on the GPU. Then a second cable to the other PCIe socket on the GPU; two entirely separate cables. Let me know how it goes. Good job on the photos, by the way. And if after that, the problem reoccurs then please open up a ticket with Corsair.

When you say about the cables, could you dumb it down for me ? I'm a bit confused as to what you mean by it

 

[rcald2000]

When you say about the cables, could you dumb it down for me ? I'm a bit confused as to what you mean by it

https://www.corsair.com/medias/sys_master/images/images/h9f/hbc/9110101786654/-CP-8920130-Gallery-CP-8920130-01.png

This cable is currently connected from your power supply to your graphics card. Currently one end of the cable is in your power supply, and the other two connectors are attached to your graphics card. What I am suggesting is that you connect only one end of your graphics card. Then get an entirely separate cable, and again, only connect one of the other GPU connectors to the graphics cards. In essence, there should be two separate cables plugged into your power supply. That's what I'm asking you to do, and then try to reproduce your system shutting off by gaming as you were before. I believe that your power supply is mistakenly shutting off due to a power protection.

 

[Minipiwi]

https://www.corsair.com/medias/sys_master/images/images/h9f/hbc/9110101786654/-CP-8920130-Gallery-CP-8920130-01.png

This cable is currently connected from your power supply to your graphics card. Currently one end of the cable is in your power supply, and the other two connectors are attached to your graphics card. What I am suggesting is that you connect only one end of your graphics card. Then get an entirely separate cable, and again, only connect one of the other GPU connectors to the graphics cards. In essence, there should be two separate cables plugged into your power supply. That's what I'm asking you to do, and then try to reproduce your system shutting off by gaming as you were before. I believe that your power supply is mistakenly shutting off due to a power protection.

I see thank you, I'll be sure to try that when I get home from work. But since we're on the topic, having two separate connections attached, how would that stop the PSU from shutting off via power protection?

 

[rcald2000]

* notes *

google query: rtx 3080 corsair shutdown

RTX 3080 causing PC shutdown while gaming. No restart, hard shutdown.
htt-ps://www.reddit.com/r/buildapc/comments/wx3u14/rtx_3080_causing_pc_shutdown_while_gaming_no/

Rtx 3080 Shutdown Issues
htt-ps://www.techpowerup.com/forums/threads/rtx-3080-shutdown-issues.291811/

RTX 3070 Ti getting shutdowns with Seasonic Focus GX-750 (Mfg. date Oct 2020) (Also with Corsair RM1000x [2021])
*htt-ps://linustechtips.com/topic/1374580-rtx-3070-ti-getting-shutdowns-with-seasonic-focus-gx-750-mfg-date-oct-2020-also-with-corsair-rm1000x-2021/page/2/

https://www.asrock.com/mb/AMD/B450M-HDV R4.0/index.asp#BIOS

I see thank you, I'll be sure to try that when I get home from work. But since we're on the topic, having two separate connections attached, how would that stop the PSU from shutting off via power protection?

I've repasted my previous notes. Please read. But more importantly, simply try it. If it doesn't work, then I suggest that you open a ticket with Corsair.

 

[rcald2000]

I see thank you, I'll be sure to try that when I get home from work.

Any update?

 

[Minipiwi]

Any update?

So sorry for not updating it, I sadly haven't had time to sit down and take my PC apart but this coming weekend I should and I'll try the recommendations. Thank you so much for the help and patience, I'll let you guys know if it works.

 

[Mark Knight]

The VRM can't handle a 5700x, that board is about as bad as it gets for am4, you'll fry your motherboard in no time, you could try putting some heatsinks on the VRM and a fan blowing directly onto it, I would just get a new motherboard personally though.

 

[Mark Knight]

AM4 Vcore VRM Ratings v1.4 (2019-11-07) - Google Drive

docs.google.com

 

[Kona45primo]

So by your logic this motherboard won't run any AM4 CPU... That's probably not correct.

 

[Avro Arrow]

Chipset driver? What is that ?

This would be your chipset driver:
AMD Chipset Driver for B450 Motherboards

I'm not sure what it does exactly but I do know that it greatly aids platform stability. I've always used AMD chipset drivers ever since my first 790FX motherboard with my Phenom II X4 940 (MSi K9A2 Platinum).