Hmmmm my CPU runs on 65C and it can still function 24/7 problem will start to occur when the CPU reaches 72C. And you did mentioned that the crashes start to occur when you enabled the sound driver so it's logical to stick to that one in order for you to solve the problem. As a suggestion if you are using your HDMI cable and going to connect it to your LCD/tv or something alike disable the onboard sound if you are just using DVI / VGA connection and a separate speaker disable your sound device from your video card. You can do this by going to your control panel or running the software that comes with your video card driver.
Network cards and sounds cards don't get along sometimes. Stupid question: I assume you've checked out device manager?
Well, technically can the 2 sound cards and the network card get a IRQ crash, something that was way more common 20 years ago but it happens.
But generally stuff like that makes the computer incredible slow but it wont bluescreen it.
I doubt it, my personal guess is that the PSU is underperforming (430W is not impressive, it should be enough but just barely. If something takes a little too much you get crashes) or heat but that are just guesses since OP havn't told us what's in his crashlog.
Now if it were an IRQ crash that shows instantly in the log.
The power supply shouldn't be a problem unless it's defective outright--which is at least plausible. He's got a 65 W CPU and a 75 W video card. This isn't a high-powered gaming rig that we're talking about. It probably doesn't often pull 150 W from the power supply. 430 W is plenty of wattage for that.
Network cards and sounds cards don't get along sometimes. Stupid question: I assume you've checked out device manager?
It's not that, sound realtek doesn't get along with win 10 , and most wireless adapter cards don't get alone with win 10, and then specifically anniversary update stalls for days with some of these, and disables your net.
I didn't mean that specifically a sound a network card don't get along, it's all WIN 10 problem with any of them separately.
It's mostly all because anniversary update doesn't play along with non-major brands of non-major products(and then some) due to device signing changes.
yes. it will only rise about 15/ 20 more c under full load. and 69/79 c is pretty normal load temps. If you get better cooling you can defenintely drop it dwn to around 35c with 50c under load
Modern CPUs might put out 2 W at idle and 60 W under a gaming load. That makes quite a difference in temperatures. The days where a CPU had all cores running at max clock speeds and voltages at idle are long over.
That said, I'm not sure if it's able to properly idle everything while in the BIOS.
bios is what regulated the voltage. so yes, the reading in bios should be accurate to the temos of all components
TDP is TDP , there is no various wattage , just because a cpu unclocks itself from a higher rate at idles doesnt change the ammount of power that is being feed into the cpu
BIOS does regulate the voltage, but software determines which P-state the processor is currently in as part of the ACPI standard. P-states vary all the way from max turbo clocks and full voltage all the way down to hibernation, with several steps in between those two extremes. P-States determine the clock frequency, voltages, and can even enable or disable parts of the CPU for power saving. The max stock clock, no turbo condition is called the P-0 state, and there are a half dozen or more that go both up (turbo clocks) and down (power saving) from there. Most people, except @Malabooga, when they talk about CPU speeds, talk about the P-0 clock speed. Intel and AMD even have brand names for their P-states - Intel SpeedStep and AMD CoolNQuiet.
Windows usually controls the P-state, it's a mixture of what you have your Energy preferences set to, and what demand is called for presently on the CPU.
The temperature is read from some on-die thermocouples, and there are usually several on a die, as well as a few installed on the motherboard in various spots. Pretty much everything that reports temperature, reports it from the hardware hooks that directly measures from the thermocouple - so yes, the BIOS will match whatever other software you have reading, because they all read from the same source.
Quiz is saying he doesn't know what P-State the BIOS has the CPU running - it could be a max-turbo condition, where the CPU would draw a good deal of power (not MAX power, since it's mostly looping NOOP commands, but still a good deal more), or a low P-state (just enough turned on to run the BIOS program itself, which would be a very low power state). Most BIOSes will tend to run in a max clock, no turbo P-0 state, which should give you a fast yet stable boot up, until your OS gets bootstrapped and can take over the P-State control of the CPU. Some motherboards I've seen you can adjust that on (ROG ones in particular).
Even at a P-0 state, a 49C idle temp is very high in my experience. I haven't ever run an FX CPU, so maybe that really is normal, but everything I see online from others tends to point to that it isn't. Even in the P-0 state, the CPU should still be down in single-digit watts, and a cooler that can handle max TDP comfortably should be more than able to keep a single-digit wattage at near-ambient temperature.
TDP is TDP - there are definitely infinite number of wattages. Power is directly proportional to clock speed, and exponentially proportional to voltage, and varies yet again based on what functions on the CPU are being utilized (a NOOP vs AVX command, for example, will draw different power requirements even at the same frequency and voltage). So power requirements vary a lot.
Even at a P-0 state, a 49C idle temp is very high in my experience. I haven't ever run an FX CPU, so maybe that really is normal, but everything I see online from others tends to point to that it isn't. Even in the P-0 state, the CPU should still be down in single-digit watts, and a cooler that can handle max TDP comfortably should be more than able to keep a single-digit wattage at near-ambient temperature.
TDP is TDP - there are definitely infinite number of wattages. Power is directly proportional to clock speed, and exponentially proportional to voltage, and varies yet again based on what functions on the CPU are being utilized (a NOOP vs AVX command, for example, will draw different power requirements even at the same frequency and voltage). So power requirements vary a lot.
Mostly good stuff, but a couple of corrections. First, it's not an FX CPU. I think it's a Kaveri with the integrated GPU disabled, but I'm not 100% sure that it's not Carrizo.
Also, power goes up fast as voltage does, but it's not exponential. I think power is proportional to clock speed times voltage squared, at least up to higher order effects. Beyond the minimum needed to run, the point of higher voltage is also to enable higher clock speeds, which makes power seem to go up faster than quadratically as the voltage rises.
To the second power isn't exponential? I'm not a math major, but my understanding was the 2 was the exponent in a "Power of 2".
No, that's polynomial. x^2 is polynomial in x, as the variable isn't in the exponent. 2^x is exponential in x, as the variable is in the exponent. The latter grows massively faster than the former:
I'm sure you've got some sort of technical background. If you've ever studied the P vs NP stuff, this is what it's about. Polynomial time algorithms (as in O(x^2)) run "fast" because polynomials grow "slowly", while exponential time (as in O(2^x)) algorithms (or more generally, anything that isn't bounded above by a polynomial) run "slow" because exponentials grow "fast".
@loke666 I dont have crash logs that I know of because its like the computer was suddenly unplugged and there isnt any crash log. Unless you can tell me where to find one I'm going to use an idea I saw on this thread that will write the log as the computer runs.
Oddly enough one of us @H0urglass experienced the same crash after updating nvidia driver. @Malabooga will love that. I'm going to see if an older driver will stabilize this sytem. I will make the crashlog thing @Wizardry told us about afterwards and hopefully that will give us some info. Then I will go through the process @13lake gave me to see if that does anything. But only after I reseat the cpu just in case.
Thank you everyone for the support and help. After many tweaks it only crashes during gaming.
@loke666 I dont have crash logs that I know of because its like the computer was suddenly unplugged and there isnt any crash log. Unless you can tell me where to find one I'm going to use an idea I saw on this thread that will write the log as the computer runs.
Oddly enough one of us @H0urglass experienced the same crash after updating nvidia driver. @Malabooga will love that. I'm going to see if an older driver will stabilize this sytem. I will make the crashlog thing @Wizardry told us about afterwards and hopefully that will give us some info. Then I will go through the process @13lake gave me to see if that does anything. But only after I reseat the cpu just in case.
Thank you everyone for the support and help. After many tweaks it only crashes during gaming.
IDK if this helps but last time it froze the power button wouldn't turn it off without holding it down. And sometimes it boots the bios 2 times before starting windows.
Velika: City of Wheels: Among the mortal races, the humans were the only one that never built cities or great empires; a curse laid upon them by their creator, Gidd, forced them to wander as nomads for twenty centuries...
Do you have a previous computer sitting around that you can swap out parts with? It's helpful to be able to selectively swap parts to see if you can trace the source of a crash. For example, if you swap the power supply and the crashes go away, then it's very likely that the power supply was the culprit. If swapping the power supply makes no difference, then you have to look elsewhere.
Do you have a previous computer sitting around that you can swap out parts with? It's helpful to be able to selectively swap parts to see if you can trace the source of a crash. For example, if you swap the power supply and the crashes go away, then it's very likely that the power supply was the culprit. If swapping the power supply makes no difference, then you have to look elsewhere.
If it's looking like it's going to cost you some money to get it running (and maybe you are at that point). You don't have parts to swap, I assume you've already pulled out all but essential parts and performed that process of elimination. I assume you've already done a full format and reinstall, then testing from the fresh install before installing each driver to make sure that isn't part of the conflict. And the memtest and P95 checks are both good tests to run as well, provided the system is stable enough to get through them.
Barring all of that, there are a couple of options when you don't have swapable parts
a) Take it into a local hardware repair shop (not Geek Squad or some big chain) - pay them to at least find the problem, if not fix it. Local geeks at a university or down the street are great if you know any, or are willing to risk Craigslisting for them, as they are often happy to be paid in beer or Redbull and love doing this kind of work, especially if they don't have to pay for the hardware themselves.
b) Start ordering suspect parts one piece at a time from Amazon or Newegg (or some place with a good return policy) - swap part out, if it fixes the problem, RMA the bad part and continue using the good one. If it doesn't fix the problem, return the new part for 15% restocking fee (or whatever it happens to be), and move on to the next.
I had one computer that ran fine for about a year. I installed Windows 10 on it. All of a sudden it started rebooting itself. Problem would kinda come and go, but at times it would get to the point where it would reboot about every 5 minutes and was essentially useless. I tried everything trying to pin down that fault - and using the Amazon method above, over the course of about 2 months had ordered enough parts to build an entire second computer. After a lot of parts juggling, I have both computers running today (son has the second one now). I still can't figure out what ultimately caused the reboots - none of the hardware tested bad, full reinstalls and rolling back to Win7 or Win8 didn't fix it after it had started, and it still has me stumped. That was the trickiest troubleshooting job I've ever had to do, and to this day, I still can't figure out what it was, because I'm still using all the original parts.
The memtest that Malabooga posted is a good check, and you should do at least one pass of that on any new build honestly, as well as some time loaded under some stress tests (I use P95 and Folding@Home right now) to make sure your cooling and power are all up to snuff before you start playing games on it and find out in the middle of a game that something is bonkers.
So I did some thinking and some more messing around following @13lake instructions. The only thing I have been able to do is make it take longer before crashing. It only crashes if one of the sound drivers is active. The onboard sound driver or the GPU sound driver. Either one of those is turned on and it crashes. If they are turned off then the computer runs flawlessly. So that leads me to 2 conclusions. Either the mobo is bad or there is a driver that conflicts.
Now if we are to assume it is a driver issue with the sound of both the mobo and GPU then that would be extremely rare. It has to be something on the mobo that is causing problems when the sound is activated. That really is the only logical explanation I can come up with. If it was anything else causing this problem then the system would crash with the sound turned off and it does not.
You're able to initially get wireless card and on-board sound working properly ?
it sounds like either the realtek drivers are making a problem, or the audio chip itself is not working properly.
Idk if it's possible to not install nvidia hdmi audio drivers, but you can try to do that, basically just install the base gpu driver to see if it crashes with realtek audio only installed from start.
Or you can try to go to realtek site and download newest version, but you have to figure out the exact name/serial key of the onboard realtek chip.
The only 3rd option is to install all the drivers as standalone, before connecting to internet/installing lan drivers. Then disable windows update for drivers, as to not let win10 get on the internet and download anything by itself. (i didnt recommend last option initially, because im not sure the tp-link is gonna work properly with standalone drivers, but its def worth a try now, as nothing else is working)
This is good advice, memory issues could be the problem.
I would try staring windows in safe mode and see if that works, safe mode wont use any drivers so if you get any issues there it wont be a software thing.
Just because it works in safe mode does not mean that the hardware is fine though, if say the PSU can't handle the power it might work in safe mode that uses less then gaming and same with heat problems.
Those are things you can do for free (and I still want to hear what your system log say). Buying additional hardware without knowing the problem might become expensive without giving you anything.
Doing a blend test in prime 95 (non-AVX version), or doing a custom test with similar settings to blend but 2000 or more RAM, and 3 min instead of 15 is also a good way to test memory.
Doing a blend test in prime 95 (non-AVX version), or doing a custom test with similar settings to blend but 2000 or more RAM, and 3 min instead of 15 is also a good way to test memory.
right version, right settings to do a separate test of either memory, clocks, voltage, overclock, stability, memory controller, etc, ...
If it was memory then it would crash if the sound was turned off. The only other reasonable explanation would be that the power is on the tipping scale and the sound draws just enough to cause the crash. That is highly unlikely as well because even with everything running 100% it won't crash with sound off. Its at a core process where the sound signals are being connected to the driver.
Comments
But generally stuff like that makes the computer incredible slow but it wont bluescreen it.
I doubt it, my personal guess is that the PSU is underperforming (430W is not impressive, it should be enough but just barely. If something takes a little too much you get crashes) or heat but that are just guesses since OP havn't told us what's in his crashlog.
Now if it were an IRQ crash that shows instantly in the log.
I didn't mean that specifically a sound a network card don't get along, it's all WIN 10 problem with any of them separately.
It's mostly all because anniversary update doesn't play along with non-major brands of non-major products(and then some) due to device signing changes.
Windows usually controls the P-state, it's a mixture of what you have your Energy preferences set to, and what demand is called for presently on the CPU.
The temperature is read from some on-die thermocouples, and there are usually several on a die, as well as a few installed on the motherboard in various spots. Pretty much everything that reports temperature, reports it from the hardware hooks that directly measures from the thermocouple - so yes, the BIOS will match whatever other software you have reading, because they all read from the same source.
Quiz is saying he doesn't know what P-State the BIOS has the CPU running - it could be a max-turbo condition, where the CPU would draw a good deal of power (not MAX power, since it's mostly looping NOOP commands, but still a good deal more), or a low P-state (just enough turned on to run the BIOS program itself, which would be a very low power state). Most BIOSes will tend to run in a max clock, no turbo P-0 state, which should give you a fast yet stable boot up, until your OS gets bootstrapped and can take over the P-State control of the CPU. Some motherboards I've seen you can adjust that on (ROG ones in particular).
Even at a P-0 state, a 49C idle temp is very high in my experience. I haven't ever run an FX CPU, so maybe that really is normal, but everything I see online from others tends to point to that it isn't. Even in the P-0 state, the CPU should still be down in single-digit watts, and a cooler that can handle max TDP comfortably should be more than able to keep a single-digit wattage at near-ambient temperature.
TDP is TDP - there are definitely infinite number of wattages. Power is directly proportional to clock speed, and exponentially proportional to voltage, and varies yet again based on what functions on the CPU are being utilized (a NOOP vs AVX command, for example, will draw different power requirements even at the same frequency and voltage). So power requirements vary a lot.
Also, power goes up fast as voltage does, but it's not exponential. I think power is proportional to clock speed times voltage squared, at least up to higher order effects. Beyond the minimum needed to run, the point of higher voltage is also to enable higher clock speeds, which makes power seem to go up faster than quadratically as the voltage rises.
x x^2 2^x
1 1 2
2 4 4
3 9 8
4 16 16
...
10 100 1024
11 121 2048
...
20 400 1048576
...
30 900 1073741824
...
50 2500 1125899906842624
I'm sure you've got some sort of technical background. If you've ever studied the P vs NP stuff, this is what it's about. Polynomial time algorithms (as in O(x^2)) run "fast" because polynomials grow "slowly", while exponential time (as in O(2^x)) algorithms (or more generally, anything that isn't bounded above by a polynomial) run "slow" because exponentials grow "fast".
Oddly enough one of us @H0urglass experienced the same crash after updating nvidia driver. @Malabooga will love that. I'm going to see if an older driver will stabilize this sytem. I will make the crashlog thing @Wizardry told us about afterwards and hopefully that will give us some info. Then I will go through the process @13lake gave me to see if that does anything. But only after I reseat the cpu just in case.
Thank you everyone for the support and help. After many tweaks it only crashes during gaming.
Velika: City of Wheels: Among the mortal races, the humans were the only one that never built cities or great empires; a curse laid upon them by their creator, Gidd, forced them to wander as nomads for twenty centuries...
2. http://www.memtest86.com/
run 2 full passes, there should be 0 errors
Barring all of that, there are a couple of options when you don't have swapable parts
a) Take it into a local hardware repair shop (not Geek Squad or some big chain) - pay them to at least find the problem, if not fix it. Local geeks at a university or down the street are great if you know any, or are willing to risk Craigslisting for them, as they are often happy to be paid in beer or Redbull and love doing this kind of work, especially if they don't have to pay for the hardware themselves.
b) Start ordering suspect parts one piece at a time from Amazon or Newegg (or some place with a good return policy) - swap part out, if it fixes the problem, RMA the bad part and continue using the good one. If it doesn't fix the problem, return the new part for 15% restocking fee (or whatever it happens to be), and move on to the next.
I had one computer that ran fine for about a year. I installed Windows 10 on it. All of a sudden it started rebooting itself. Problem would kinda come and go, but at times it would get to the point where it would reboot about every 5 minutes and was essentially useless. I tried everything trying to pin down that fault - and using the Amazon method above, over the course of about 2 months had ordered enough parts to build an entire second computer. After a lot of parts juggling, I have both computers running today (son has the second one now). I still can't figure out what ultimately caused the reboots - none of the hardware tested bad, full reinstalls and rolling back to Win7 or Win8 didn't fix it after it had started, and it still has me stumped. That was the trickiest troubleshooting job I've ever had to do, and to this day, I still can't figure out what it was, because I'm still using all the original parts.
The memtest that Malabooga posted is a good check, and you should do at least one pass of that on any new build honestly, as well as some time loaded under some stress tests (I use P95 and Folding@Home right now) to make sure your cooling and power are all up to snuff before you start playing games on it and find out in the middle of a game that something is bonkers.
Now if we are to assume it is a driver issue with the sound of both the mobo and GPU then that would be extremely rare. It has to be something on the mobo that is causing problems when the sound is activated. That really is the only logical explanation I can come up with. If it was anything else causing this problem then the system would crash with the sound turned off and it does not.
it sounds like either the realtek drivers are making a problem, or the audio chip itself is not working properly.
Idk if it's possible to not install nvidia hdmi audio drivers, but you can try to do that, basically just install the base gpu driver to see if it crashes with realtek audio only installed from start.
Or you can try to go to realtek site and download newest version, but you have to figure out the exact name/serial key of the onboard realtek chip.
The only 3rd option is to install all the drivers as standalone, before connecting to internet/installing lan drivers. Then disable windows update for drivers, as to not let win10 get on the internet and download anything by itself. (i didnt recommend last option initially, because im not sure the tp-link is gonna work properly with standalone drivers, but its def worth a try now, as nothing else is working)
I would try staring windows in safe mode and see if that works, safe mode wont use any drivers so if you get any issues there it wont be a software thing.
Just because it works in safe mode does not mean that the hardware is fine though, if say the PSU can't handle the power it might work in safe mode that uses less then gaming and same with heat problems.
Those are things you can do for free (and I still want to hear what your system log say). Buying additional hardware without knowing the problem might become expensive without giving you anything.
This is an awesome link to get started :http://overclocking.guide/stability-testing-with-prime-95/
right version, right settings to do a separate test of either memory, clocks, voltage, overclock, stability, memory controller, etc, ...