I tried running a stress test on the cpu and gpu for 8hours and it works fine its only when i remove the molex power cable from the motherboard the system freezes and the diagnostic led lights up.. it was never like that before

Hi Samson and welcome to the R.O.G forum.

Have you tested a different power supply at all as the CX 430 is not enough to power an oc'd 8350 and a GTX 680 when the 680 requires about 550w. According to your psu's spec's it only has 1x 6+2 PCIE so what cables do you have connected to the gpu's power connection?

http://www.guru3d.com/articles_pages/geforce_gtx_680_review,9.html

The CX430 V2 only has a single 28A-rated 12V rail with which it can output a maximum of 336W; at this price, this amount is plenty. The PSU is also rated to produce 20A from both its 3.3V and 5V rails, and 3A from its 5VSB rail.

Quoted from this review http://www.bit-tech.net/hardware/2011/09/13/what-is-the-best-400-599w-psu/4

Thanks for your reply and no i haven't tested the system with a new smps yet but i'm sure the smps has enough juice to run the graphic card without problems i'm using the 1x 6+2 PCIE and a 4 Pin Molex to 6 Pin PCI-Express Convertor Cable to power the graphic card it was working fine 24*7 for 6 months (i do mine alternate coins) and its working fine now but freezes as soon as i remove the molex power connector from the motherboard that is designed to give extra power to the pci-e slots

First the EZPLUG (molex plug near the PCIe slot nearest the CPU socket) has been implemented on this motherboard to ensure that enough wattage is supplied to the PCIe slots (the standard is that each PCIx16 slot is able to provide 75 watts of power). I know according to the manual this is only needed when using two graphic cards but if you are having the VGA LED on then your graphics cards is pulling a lot of extra power so be sure to keep the EZPLUG populated.

Secondly as already stated you need to upgrade your PSU for the following reason:
a) you don't have enough power plugs to properly power your CPU let alone your Graphic Card. The use of "cheater cables" do nothing more than add extra load to a already overload "rail" of the PSU.

b) Also burning your system is the worst you can do because if you connected a PWM (Power Watt Meter) you would see that your system draws about 140 watts doing nothing and when you stress test your system you are using well over 400 watts (I have a Matrix R9 280X plus a FX-8350 and I draw well over 550 Watts when stress testing).

With that said I would start with a new PSU...I would suggest the Thermaltake DPS series PSU as they have a USB port where you can monitor the wattage and voltages directly from the PSU).

I would also reset your memory and maybe try only running one memory module at a time to see if you can get past the DIMM_LED error.

Meantime makes sure you have the following connectors populated:
-Aux 4-Pin CPU Power
-Aux 8-Pin CPU Power
-EZPlug
- All PCIe connectors on your graphics card

Post back with more questions.

The issue was resolved after a bios flash. [BIOS V 2101 2014/05/22]

Thank you for posting back with the outcome.

Enjoy your system