Stops working for no reason
D:\oclHashcat-plus-0.09>oclHashcat-plus64.exe -m 0 -a 3 -n 600 --gpu-loops 1024 -o d:\7h-main.txt --outfile-format=2 --remove d:\MD5h.txt -2 ?l?u?d!@#*-_. ?2?2?2?2?2?2?2
oclHashcat-plus v0.09 by atom starting...

Hashes: 8716449 total, 1 unique salts, 8716449 unique digests
Bitmaps: 21 bits, 1048576 entries, 0x000fffff mask, 4194304 bytes
Workload: 1024 loops, 600 accel
Watchdog: Temperature abort trigger set to 90c
Watchdog: Temperature retain trigger set to 80c
Device #1: Cypress, 1024MB, 765Mhz, 18MCU
Device #1: Kernel ./kernels/4098/m0000_a3.Cypress_938.2_1.4.1741.kernel (568652 bytes)

[s]tatus [p]ause [r]esume [b]ypass [q]uit => s
Status.......: Running
Input.Mode...: Mask (?2?2?2?2?2?2?2)
Hash.Target..: File (d:\MD5h.txt)
Hash.Type....: MD5
Time.Running.: 2 mins, 10 secs
Time.Left....: 4 hours, 59 mins
Time.Util....: 130869.7ms/276.9ms Real/CPU, 0.2% idle
Speed........:   411.0M c/s Real,   411.9M c/s GPU
Recovered....: 0/8716449 Digests, 0/1 Salts
Progress.....: 53791948800/7446353252589 (0.72%)
Rejected.....: 0/53791948800 (0.00%)
HWMon.GPU.#1.: 99% Util, 51c Temp, 47% Fan
[s]tatus [p]ause [r]esume [b]ypass [q]uit => s
Status.......: Running
Input.Mode...: Mask (?2?2?2?2?2?2?2)
Hash.Target..: File (d:\MD5h.txt)
Hash.Type....: MD5
Time.Running.: 9 hours, 40 mins
Time.Left....: 21 hours, 6 mins
Time.Util....: 33846185.1ms/291.2ms Real/CPU, 0.0% idle
Speed........: 67806.4k c/s Real,   424.0M c/s GPU
Recovered....: 35856/8716449 Digests, 0/1 Salts
Progress.....: 2294988595200/7446353252589 (30.82%)
Rejected.....: 0/2294988595200 (0.00%)
HWMon.GPU.#1.:  0% Util, 40c Temp, 45% Fan
[s]tatus [p]ause [r]esume [b]ypass [q]uit => q

GPU activity 0%. Speed shows 67K is incorrect, cause after I press status progress is the same. Yesterday was the same thing. Gets stuck around 30%. Doesn't quit properly after, need to use ctrl+break. Before I leave its working fine with ~410c/s. This MD5h.txt I been working on for a while with no problems. Cracked 25% so far. But now freezes with this particular mask. What the hell.

Win7.64, 8GB, Catalyst 12.8, Plenty HDD space.
Very strange thing, sounds like hardware problem
Using default --gpu-accel --gpu-loops finished with no problems. Is any more reporting on this issue necessary? Or its a purely hardware fault. I guess no point looking into as this only affects me.
On my setups it happens because of:
1. PSU, too low wattage to feed gpu's on higher clocks.
2. ASIC hangs, when GPU too much overclocked to be busy at 100%
Thanks for replies. Got Corsair AX850, which gets good reviews. I measured whole PC's power consumption on load - 280W. And I run default clocks for my HD5850.

My CPU is overclocked. Q6600 2.4 to 3.2ghz for 4 years now. Could that be a problem? Tested with LinX, no faults.