Computer shutdown after running Cgminer for 40-60 minutes

[EarlGrievous]

Truly confused as to why my Computer is shutting down after running cgminer for 40-60 minutes.  Any advice is much appreciated!

Computer runs perfect with no shutdowns while not running Cgminer.

Cgminer starts up perfectly fine and is running stable with no HW errors or any signs of complications.  Running a single Sapphire R9 280x with a 1200Watt PSU.  My first assumption was a lack of power running into my GPU and I've tried all three PCI power connections (but with the 1200Watt PSU - power shouldn't be an issue).  Running SDK 2.9 with 13.11 AMD driver.  I've tried multiple cg conf files as well (current conf file shown below).

Temp avg: 74C
GPU Load: 99%
Fan speed: 40%
GPU Core Clock: 1080 Mhz
GPU Memory Clock:  1500 Mhz


Hardware listed below:
1)  MB GBT|GA-990FXA-UD3 AM3+ (Rev.4.0)
2)  VGA SAPPHIRE|100363L R9 280X 3GD5 R
3)  PSU SST | SST-ST1200-G EVOLUTION R
4)  MEM 4Gx2|G.SKILL F3-12800CL9D-8GBXL
5)  CPU AMD|4-CORE FX-4130 3.8G 4M R


{
"pools" : [

{
        "url" : "middlecoin.com:3333",
        "user" : "wallet address",
        "pass" : "xx"
}

   ]
,


"intensity" : "13",
"vectors" : "1",
"worksize" : "256",
"kernel" : "scrypt",
"lookup-gap" : "2",
"thread-concurrency" : "8192",
"shaders" : "2048",
"auto-fan" : true,
"temp-target" : "75",
"expiry" : "120",
"gpu-dyninterval" : "7",
"gpu-platform" : "0",
"gpu-threads" : "2",
"gpu-engine" : "1080,1080,1080,1080",
"gpu-memclock" : "1500,1500,1500,1500",
"gpu-powertune" : "0",
"log" : "5",
"no-pool-disable" : true,
"queue" : "0",
"scan-time" : "30",
"scrypt" : true,
"shares" : "0",
"kernel-path" : "/usr/local/bin"

}


Will make BTC donation for anyone who can solve this!! 

[Starscream]

Despite what you wrote, I'd still venture to say it's the temperature of your GPU that causes that shutdown. What's your VRAM temperature close to the 30-40 minutes mark?

Generally speaking, a computer will shut itself down to protect a piece of hardware, usually the cause is overheating.

[LetsTalkBitcoin]

Id say temperature is your best guess. Not necessarily the CPU but basically any piece of hardware with a failsafe mechanism. Check CPU, RAM, NorthbridgeSouthbridge Chips, PSU, etc.

• What if you open the sides of the case? Does it run longer?
• Did you try ramping up the fan speed from 40% to something like 60% or 80%?
• Maybe set "temp-target" to a little less like "70"?

[pengoau]

check your vrm temps with gpuz, they shouldn't be exceeding 100C

[robelneo]

Cpuid is a must if you want to mine coins I have installed on all the machine where I am mining coins it gave me detailed analysis of the temperature of my pc

[Equate]

which cgminer version are you using.Whats is your OS.

[EarlGrievous]

Replying to all response here.

1)  Running 3.7.2 Cgminer with windows 7 64-bit

2)  I use GPU-Z to monitor while mining and doesn't report VRM temp.... how do I check this?

3)  Giving Cpuid a try

4)  Both sides of my case are already off - 74C doesn't seem excessively hot for a 280x from what I've been reading on forums....

I'll try setting temp-target to a lower setting and ramp up fan speed as well

Furthermore, I'm trying to switch out RAM from 8 to 16GB.  If that doesn't work trying a faster CPU next.


Will report back after I try there steps.....

[Starscream]

In GPU-Z you should be able to see VRAM temperature, it will be at the bottom of the sensors list (you can scroll down if you haven't noticed).
It also wouldn't hurt to monitor the temperature of your CPU.

And you really don't need more RAM or a faster CPU - speed isn't your issue here.

[carly200]

I did not check the previous posts, just your headline...

If you have a multi rail PSU and overload a rail, it may shutdown (like a fuse) your system.

that may happen after some time, when the systems (GPU) has heated up and draws more power...

[Wipeout2097]

If you have a multi rail PSU and overload a rail, it may shutdown (like a fuse) your system.

that may happen after some time, when the systems (GPU) has heated up and draws more power..

This.

@OP, if Gpu-z doesn't show VRM temp, try HwInfo

[EarlGrievous]

I have a single +12V rail so don't believe that's the cause.

Tried changing "Temp-targ" to 70 and still shutdown..... trying 65 now.  Upped the fan speed to 65% too.... still not the case.

GPUZ does not list "VRAM Temp" for the 280x under sensors.  Downloaded Hwinfo and still don't see "VRAM Temp" anywhere.  Do some GPUs not show VRAM Temp or something? 

CPU avg temp = 8.3C
Motherboard avg temp = 30.9C
Northbridge avg temp = 31.6C
VR T1 avg temp = 40C     (don't know what this is?)
VR T2 avg temp = 40.2C   (don't know what this is?)
GPU avg temp = 64C  (with new temp-target set to "65")

[Starscream]

I'd venture to guess that VR is your VRAM.

A stupid thought, we sure your PC doesn't go to sleep? go to control panel, go to hardware and sound, go to power options, select change when the computer goes to sleep (never), win?

[EarlGrievous]

Not a stupid thought.... but yes I've changed both Sleep and Monitor settings to "never" turn off.

[miner765]

Are you living in a house or apartment? Try switching power outlet, what PSU do you have?

[serje]

I recommend you to put the pc in a room where it's cold since it's still winter!
Open windows and don't enter that room

Of course if you have a spare room otherwise just use your balcony or terrace if you have one ... just be sure rain won't get to him and he has enough air to "breath"

[EarlGrievous]

PSU
PSU SST | SST-ST1200-G EVOLUTION R

I live in an apartment, and trying to switch power outlets.  If that doesn't work I'm willing to put it on my balcony to test it..... just want this to work :/

[Prelektro]

My bet it's vrm temperatures %99. If computer is shutting down, it means something is really getting hot and its either cpu or gpu. New stock Sapphire cards have serious problems about vrm temperatures, month ago my 4 sapphire dual-x 7950 got burned and customer service said there were melted solderings. I heard sapphire r9 280x's also have that problem. Check your bios number, if it's 015.041 that bios has vrm temperature issues. You can flash your bios with different one but be careful it may cancel your warranty.

[EarlGrievous]

Bios Information:  113-5E2490U-O46 for the R9 280x so doesn't look like it fits the below description of 015.041

[Prelektro]

It's not bios version.

Check gpu-z , here is my sapphire r9 280x toxic bios. I got 015.039 and it's working without an issue.

[EarlGrievous]

Oh shit.... yeah do have Bios version 015.041...... but switched to an outlet on the complete opposite side of my apartment and cgminer has been running for over an hour now.  I may have been pulling too much power from one outlet before (had another cpu running on that outlet plus surround sound and a tv).  Hoping this holds while I grab a couple drinks