New Computer - "Black Screen" - Fan "Ramps Up"

Hello All,

I just received a brand new computer a few days ago from CyberPower PC. I've attached the specs below.

That being said, I noticed a problem when I was "in a game" and within an hr, the fan revved up, and the computer shut down to black screen (monitor read "no input"). Thinking this was a one-off, I tried again, and the same symptoms persist - after playing for some time, the screen goes blank (no error messages), the fan rev's up, and requires a hard reboot. As an FYI, I paid Cyberpower to OC this computer for me. No error message whatsoever.

Next, during one of the "hard reboots", a message popped up before windows that said the setting were unstable, and that I had to reduce my clock speed down - so I clicked accept (there were no other options to potentially ignore this message).

Here are the steps I've taken to date:
1. Upgrade the MOBO / GPU drivers.
2. Plugged the system directly into the outlet to ensure adequate power (as opposed to power bar).
3. Ensured RAM and GPU was seated properly.
4. Ran MEMTEST - no issues.
5. Ran Prime 95 -> Torture Test -> Blend - Causes a "FATAL ERROR" every time as per below:

"FATAL ERROR: Rounding was 0.5, expected less than 0.4"
"Hardware failure detected - consult stress.txt file"
Torture Test Completed 181 Tests in 2 hrs 57 minutes - 1 error, 0 warnings" .

Each "worker" in this test (I'm assuming for each core) had a different result as per below:

Worker 1 - no error
Worker 2 - Rounding was 0.5, expected less than 4
Worker 3 - Rounding was 0.5, expected less than 4
Worker 4 - Rounding was 0.4996338901, expected less than 4.

My CPU spec's are as follows:


CPU: Intel® Core™ i5-3570K 3.40 GHz 6MB Intel Smart Cache LGA1155 (All Venom OC Certified)
CD: 24X Double Layer Dual Format DVD+-R/+-RW + CD-R/RW Drive (BLACK COLOR)
CAS: Cooler Master HAF 912 Mid-Tower Gaming Case w/ Adjustable HDD Cage (Black Color) [+18]
FREEBIE_MB: FREE! 802.11N Dual Band Wireless + Bluetooth 4.0 Expansion Card [+0]
FAN: Asetek 510LC Liquid Cooling System 120MM Radiator & Fan (Enhanced Cooling Performance + Extreme Silent at 20dBA) (Dual Standard 120MM Fans (Push-Pull) [+9])
HDD2: 1TB Western Digital Caviar Blue SATA-III 6.0Gb/s 7200 RPM HDD [+94] (Single Drive)
HDD: 250 GB SAMSUNG 840 Series SATA-III 6.0Gb/s SSD - 540MB/s Read & 250MB/s Write [+107] (Single Drive)
IUSB: Built-in USB 2.0 Ports
MOTHERBOARD: * [CrossFireX/SLI] GIGABYTE GA-Z77X-D3H Intel Z77 Chipset DDR3 ATX Mainboard w/ IRST, Lucid Virtu MVP, Ultra Durable4 Classic, 7.1 HD Audio, GbLAN, 3x Gen3 PCIe x16, 3x PCIe x1 & 1 PCI (All Venom OC Certified) [+41]
MEMORY: 8GB (4GBx2) DDR3/1600MHz Dual Channel Memory (Patriot Viper 3 Intel Extreme Masters Edition [+9])
NETWORK: Onboard Gigabit LAN Network
OVERCLOCK: Extreme OC (Extreme Overclock 20% or more) [+49]
OS: Microsoft® Windows 7 Home Premium [+0] (64-bit Edition)
POWERSUPPLY: * 750 Watts - Corsair CMPSU-750TXV2 80 Plus Power Supply - Quad SLI Ready [+57]
SOUND: HIGH DEFINITION ON-BOARD 7.1 AUDIO
TUNING: Intel® Core™ i5-3570K Performance Tuning Protection Plan by Intel [+24]
VIDEO: AMD Radeon HD 7950 3GB 16X PCIe 3.0 Video Card [+187] (Major Brand Powered by AMD)

FYI - I've run Speedfan during the crashes in Prime95 and during gaming, and no temperatures seem off (the highest the CPU got was ~ 50 Celsius).

The system is under warrranty - but this isn't the ideal scenario for me as the shipping to fix it will cost me approx $200.

Andrew.

Seems like a thermal shutdown to me. When CPU/GPU overheats, the system either reduces their clock in order to cool them down or does a hard shutdown to protect the hardware. You said you OCed your hardware. Make sure that your are running the CPU/GPU at stock clocks and not overcloking them and also make sure the cooling system is working properly (ie, the fans) and air is flowing good. Thanks for reading.

HI Roboticarm,

Yes - CyberPower OC'd the computer for me. That being said, I did reset this (i.e// went into BIOS and chose "restore optimized defaults" and it's no longer OC'd. I verified this (i think) by running Prime95 and looked at the speed - never got above 3.4GHZ. I have never done anything to the GPU. In fact, I ran furmark for a good while last night and the highest temp the GPU got to was about 70C.

The only time i can reproduce the crash is a) in the game, or b) when running a Prime95 blending test it occasionally crashes the system.

Regarding the cooling, I've had the case open and looked at the fans - they "seem" to be working. It's cool like a cucumber in there... I have 2 fans (push/pull config) and also liquid cooling. I ran Prime95 / Furmark at the same time yestrerday for 30 minutes to try to stress the system out, and the CPU never got about 55-60C or so.

Any thoughts?

Andrew.

Hello!
Just wanted to update this thread - i contacted CyberPower and they provided an RMA number for me to return the system - that being said they believed the problem was caused by the Video Card (given the symptons of the black screen). They suggested i tried using the onboard video. So here's what i did:
I uninstalled the AMD 7950 card and all drivers (used driver sweeper after). I then physically removed the card and rebooted with onboard video. All symptons now seem to be resolved. I.E.// the games that were crashing never crash, and i just ran prime95 on blend for 12 hrs with NO errors. Could of the video card cause the crashes in the game/prime95 errors? Without the video card, everything run's smooth as can be.
What i did notice is that even though i have an AMD video card, when looking at driver sweeper, i had some Nvidia Drivers (Display, and 2 others) - can i remove these? Not sure why these would have been installed? APpreciate any advice going forward.
I'm going to try to do another clean install of the video card and see if it works now (Error free). If the errors come back - can i assume that i should be RMA'ing just the video card?
Thanks,
Andrew.