Alienware 17 - Immediate Shutdown issues

Hi all,

Posting this as I cannot find any answer anywhere, and am nigh on running out of ways to solve it.

This week, Monday, my laptop (Alienware 17, i7, 880m), suddenly cut out during gameplay. Thinking it might have been a software issue, I rebooted the machine, and tried to diagnose the error.

Since then, the laptop has been turning itself off, predominantly when the GPU is under load/ stress. It should be noted that the max temperatures reached are 65C for the CPU, and 74C for the GPU - Thermal paste has been reapplied, all vents/ fans cleaned, and both fans working (indeed, they speed up when required). 

The main times the laptop turns off is during Furmark stress testing, when playing a game (for testing since the issue), and sometimes when the laptop is rebooted following the sudden power loss. Kernel-41 issues are recorded each and every time.

Additionally, I have:

- Done a clean install of 8.1 AND a clean install of windows 10

- Clean installed/ changed/ rolled back/ updated graphics drivers and all other drivers necessary

- Flashed BIOS

- Replaced motherboard battery

- Ran tests on no battery and only power lead

- Ran tests on battery and no power lead

- Disabled devices to balance power and check for issues

- Ran multiple stress programs - errors only appear when connected to the GPU (during stress testing or following stress testing/ a shut down).

- CPU has been stressed and no issues caused

- Laptop is just over a year old (1 year and 10 days since I received it).

Any suggestions as to causes/ fixes? I've a nagging suspicion it's a PSU issue, as it only occurs under load, rather than immediately trying to engage hardware (GPU burnout would mean it would be unresponsive, undetectable and cause shut down immediately on being engaged - but this can be 15seconds - 2minutes after it's being utilised).

Bit disappointed, as I've loved the laptop since I've got it, and it's only just over a year since I've received it. [also, after an absolutely awful 20days, I wouldn't mind playing some games to rest and relax... but that's nigh on impossible now].

Further update: Ran without battery on PSU, cut out upon launching a game after 10 seconds.

Ran on battery at maximum settings for GPU in power management, managed a good 20minutes play without cut out. 

However, should I trust the sensor readings from say speccy? Monitoring the temperature, it reached a maximum of 58C - however, touching the heat sink (after it spontaneously shut down), it seemed to be hotter than this - even with (completely clean/ clear) fan buzzing away (Nevertheless, I understand the process of the heat sink is to draw heat away.. and so will, in all likelihood, get hotter than the GPU).

Additionally, this seems to be my only avenue for finding SOME resolution. My calls this week have gone unanswered, or I'm simply given flat responses, and no replies to emails either. So, any help is much appreciated!