Ok how are you stressing the hardware and what software are you using to monitor clocks and temps?
Use CPU-Z to make sure your CPU really is stock 3.4GHz, and use Prime95 to stress test for stability. Let this run for 10 minutes while you monitor temps with HWMon or Coretemp (standalone).
For GPU use MSI kombustor to stress test the GPU and monitor the temps.
If either individually don't shut down the computer then run them both together. If that shuts down your PSU isn't providing enough voltage.
The guy barely has a 300w load, he doesn't need 800+ watts for such a minimal system.What case do you have? It could be as simple as an ventilation problem. You need to have more CFM going out than coming in. This creates a negative air pressure environment. It aids in pulling air in to the case taking the load off the intake fans.
I just rebuilt my system using an Antec P280 case. Although it can support intake and internal fans, they turned out to impede the flow of air in to the case. After removing them and doing a through clean up including the heat sink, it runs cooler and quieter.
The CX430 is not enough to push all the devices you have and do a CPU over clock that is stable. I'd be upgrading that PSU to at least and AX850 or AX860. They are completely modular. I like mine. I just keep adding devices and it doesn't even work up a sweat.
CPU is 100w TDP and he isn't using the GPU portion which means he should be pulling less than 100w from the PSU under normal CPU usage.CPU Maximum power draw is about 250w AFAIK, GPU can pull upwards of 130w... So yeah, those two components alone if this happens only during load could be stressing the PSU too much once you add in all the other components...
Have you done as PP said earlier with those stress tests to see if the machine powers down?
Yea but did you do exactly as I said? What were the results? Did it shut down running one or the other or both at the same time?Thank you for the idea or whatever, so People were sadly rigth and it was my PSU that was a problem and i am now on the look for better PSUs.
Thanks to everyone Helping me find the solution!
Ok how are you stressing the hardware and what software are you using to monitor clocks and temps?
Use CPU-Z to make sure your CPU really is stock 3.4GHz, and use Prime95 to stress test for stability. Let this run for 10 minutes while you monitor temps with HWMon or Coretemp (standalone).
For GPU use MSI kombustor to stress test the GPU and monitor the temps.
If either individually don't shut down the computer then run them both together. If that shuts down your PSU isn't providing enough voltage.
TDP isn't total draw, it's thermal design point... 100w is the total heat the CPU can put out at maximum load, you can't put out more heat than what you pull in... Also, that CPU draws like 50w at idle minimum if my memory serves right.
Anytime you have a TDP, the actual wattage pull of the CPU will be higher than the TDP under load.
You contradicted yourself right there. AMD CPUs are inefficient but damn, you were trying to say this thing pulls 250w, sorry, lack of sleep? not sure how I contradicted my self, I was basically saying, you can't put out 100w worth of heat unless you are pulling in more than 100w worth of power.. No, that would be closer to the 9570, not a little APU. This review states a 130W full load, which granted they don't tell how they measure I'm going to guess it's full system by a kill-a-watt meter. (This is why I take ALL power consumption values from TPU If you are talking about TechPowerUp, they don't even show a power consumption for the 750 or 750k, just saying, I will stick with Toms as far as power consumption goes, till someone wants to get a socket extension and actually do real engineering.). Given values, actual CPU pull should be closer to 100w since the idle state of the GPU should be around 15w that they used, plus HDD.
In any case, let's say that 130W total power consumption was the actual value of the CPU if we add that to the 130W of the GPU it only consumes half of the rated wattage of the CX430. Meaning in all reality his little system isn't pulling anything the PSU can't handle. I think he has a bad unit with his results.
Was that meant for me or for C0rr0sive?Actually you can, go talk to an electrical engineer. Not even going to bother with that convo because I already proved that his PSU should be fine in theory, which is the subject of the matter.
TPU has better equipment to measure than any other site on the net. They have the best PSU testing equipment and the best methodology of measuring power consumption from the device itself. For instance every single GPU reviewed has the power consumption measured at the power input rather than a box on the wall. It's why I link specifically to them, and obviously they didn't have a 750k which is why I linked the other site that clearly measures the way all other sites measure, from the wall. If you want to take numbers from Tom's then feel free not to regurgitate their BS here.