Rpoblems with "random" gaming crashes

[brun0]

Hello all,

Thank you in advance for reading this long post! I appreciate it!

I have been searching the internet for over a week now and just when I might have found a forum in which a question similar enough to mine has been asked, it never gets answered, so I hope by asking it here specific to my system I could have some luck

I recently built my gaming computer and everything about it is great, except for one issue. When I am playing what I would call system intensive games (so far DayZ and Guild Wars 2), at some point the system will crash. I play them on ultra settings with everything except for AA on.

It could be an hour in, it could be 5 hours in; it honestly feels very random. The crash is as follows: Monitor goes black and a message displays saying "no signal input", the sound goes all weird (if using a headset, otherwise no sound since it would go through HDMI cable onto signal-less monitor). Then I have to force restart computer holding down power button. Upon restart it shows the "Windows didn't shut down correctly, chose safe mode or start normally" window. Once I click restart normally, everything is fine and back to normal until next intense gaming session.

I have changed HDMI cables, I have changed PSU (I will get to this in a minute). Catalyst Drivers are up to date (12.8) and even in Sapphire Trixx it reads the drivers as 12.8, so it reassures me that they are installed correctly. I have reseated the GPU just in case. No dirt, a lot of airflow (modular PSU and cables tucked behind towards the back plate of case).

This is my build:

CPU: i5-2500k OC'd to 4.5
CPU Fan: Cooler Master 212 EVO
GPU: Sapphire Radeon HD 7870 2GB GHz OC Edition
MoBo: ASUS P8Z77V-LK
Hard Drive: Seagate Barracuda 1TB 7200
RAM: Corsair Vengeance 8 GB (4x2) 1600
CASE: Corsair 500R Mid Tower
PSU: 850W Bronze Plus Thermaltake ***(Had Corsair TX 650W Bronze Plus but switched it to 850 as a recommendation from the Master guy at Geek Squad computer center)
Optic Drive: a Sony drive, nothing fancy.
Monitor: a 19'' Samsung HDTV (resolution set to 1360x768 to see if bringing it down might help, it didn't) --> since it is a TV system uses generic drivers)

I have logged GPU-Z and the temps do not go above 60 when the game crashes. I attempted to log HWMonitor using the Pro Trial but some information is missing. Nonetheless I think a lot of it is useful to hopefully figure out the problem(s).

Since it was not the PSU (and the PSU is single rail so that's a good thing right?), I have come to think the following:

*I have the OC'd version of the GPU, so maybe it is unstable? Does not seem likely since GPU-Z readings do not seem weird (I can re-do a log and after it crashes post that one on here just to double check)
*GPU drivers, but two different programs read them as the correct Catalyst Drivers (12.8)... unless the older drivers are still messing with the system?
*My 8 GB of RAM are not working too well, but the BIOS reads 8000+ MB RAM (something like 8,256)... with voltage at 1.5. Have memTest installed in case I should run more tests.
*The CPU might be pushed during intense gaming sessions, but wouldn't the screen go blue instead of the monitor just lose signal?
*The HDTV is just no good for gaming, and at a 60Hz refresh rate maybe something happens where the generic drivers just don't function at some point and the computer cannot identify the monitor, but it should come back when i turn the monitor off and then on?

I am very confused, the Geek Squad man seemed very confident when he said it was the PSU, but now I am stumped.

I genuinely appreciate the help.

Thanks in advance,
Bruno

 

[brun0]

I should also note that I had some trouble updating my BIOS. I initially used AsusSetup to update it, and during the installation all went well and it told me to "reboot to finish installation". When the system was rebooted, it gave the 2 following messages:

"No Such File Found
Can't find AsusSetup.ini :
C:\Windows\Intel_Chipset_V9301019_LK_XPWin7\Driver\hotfix\AsusSetup.ini"

The other along the same lines, but in D: instead of C:, and it happens only when I have my USB plugged in as I start windows.
(more on this second part in a sec)

So it could not finish the installation. I then decided to go back to the ASUS website, download the .cap file for my MoBo onto my USB drive and use Asus EZ Flash to update. The update was successful this time, but now every time I log in it shows that aforementioned message. (the second message when this USB is plugged in).

Could this be the problem? The update was done successfully, so I don't think it is; I just thought it was worth mentioning.

 

[chusteczka]

First thought would be not enough power to the system but your 650W and definitely the 850W PSU's should be fine. You may want to check your system with a Kill a Watt load meter just in case. However, 600W of power is insane and I doubt your system or video card is using that much.

If it is not the power, then I suspect the video card. Sapphire Tech is a reliable brand of video cards. I believe it is the original AMD video card provider. If you do not get a proper answer here, try the Sapphire Tech forums for better help.

It is expensive to swap out video cards. For this reason, it might be helpful to go directly to the Sapphire forums for their expert assistance. They should be able to tell you quickly if the symptoms point to a failed card.

Is it possible to decrease the overclock speed on your video card? Maybe try decreasing by 10% - 20% and see if that helps. If it does, then you have solid evidence to return your video card.

Since you mention the bios, it might help to double check any video bios settings.

I just noticed that your Intel Core i5 cpu is overclocked to 4.5GHz. Now your symptoms make sense. Typically, an unstable video card will provide crazy colored graphics as it fails during a crash. The immediate loss of video signal with "no signal input" on the monitor is typical to cpu overclock instability.

Drop your cpu overclock down to a more stable setting and run your system as normal. I bet you will not crash with a decreased cpu overclock.

An alternative test would be to decrease the cpu overclock to 4.0 or 4.2GHz and also decrease the video card overclock at the same time. That should definitely solve this issue. Then slowly increase one or both and test for stability.

Prime95 or other cpu overclocking system stability tests tend to focus on floating point computations and memory access. Your cpu could very well be stable in these areas but may then be unstable with an extreme video card adding the strain of northbridge signals between the cpu and the video card. In fact, check your northbridge voltage. This may need to be increased a little to handle the combined load. Not enough northbridge voltage would then reconnect with initial thoughts of the power supply not providing enough electricity.

We have now come full circle and I feel more confident that your northbridge needs more voltage. Be careful with adjusting your settings. Think everything through carefully so you do not put too much voltage through the system. If necessary, drop your cpu overclock a bit if your voltages are too high.


Further searching shows the Intel Z77 chipset does not have a northbridge. The video is directly connected to the cpu. I have not kept up with Intel advancements so missed this. Is there a voltage setting you can use for video signal communication? Possibly a PCI-E video voltage setting.

newgg.com - ASUS P8Z77-V LK LGA 1155 Intel Z77 HDMI SATA 6Gb/s USB 3.0 ATX Intel Motherboard with UEFI BIOS

http://www.ocfreaks.com/asus-p8z77-v-deluxe-review/

Read up on these reviews. You may find something to help.
anandtech.com - Intel Z77 Motherboard Review with Ivy Bridge - ASRock, ASUS, Gigabyte, and MSI
tomshardware.com - P8Z77-V Pro Firmware
ocfreaks.com - ASUS P8Z77-V Deluxe Review
techspot.com - Intel Z77 'Panther Point' Chipset Overview
google.com - z77 northbridge


I assume you have both 6-pin PSU connectors connected to the video card.
newegg.com - SAPPHIRE Radeon HD 7870 GHz Edition OC 2GB 256-bit GDDR5 PCI Express 3.0 x16 HDCP Ready CrossFireX Support Video Card (11199-03-20G)

 

[Puppies04]

^ This, however the most important part is this.

Do not try to fault find while overclocked ( I would consider a factory GPU OC acceptible)

Other things to check if problem persists, in no particular order.

*Reset mobo by removing the CMOS and holding the case power button for 30 secs while the machine is unplugged from the wall

*Make sure HDD SATA lead is secure

*Change HDD SATA lead

*Drop ram down to 1333mhz (I have come across a couple of sets of corsair "1600mhz" RAM that wasn't completly stable at the rated speed)
*Make sure you aren't getting voltage spikes/dips from the wall outlet (remove anything else on the same ring main that turns on/off periodically and draws a large current/try another wall outlet/get a UPS)

*Make sure the mobo 20+4 and 8pin power connectors are pushed in properly

*Run some HDD diagnostics to make sure the drive is operating normally.

Hope you manage to find what is wrong, post back with results.

 

[brun0]

chusteczka,

I am so happy to read your reply!

I will get home at roughly 11pm tonight (am in community college from morning until then).
I actually reconfigured the OC setting to see if that was the case, ran prime95 for almost 6 hours at 4.5 and nothing went wrong, but I will clock it down and try it then.
What I will also do is take a screen shot of every page in the bios using my USB drive and put every screenshot on here, so you can see exactly what I am seeing.

I also connect the HDMI directly to the video card.. when I ahve tried to connect it to the back of the MoBo. I believe there has been no signal, which is also confusing.

Out of the "optional" in your z77 screen shot, I believe I have:
-intel rapid start
and maybe intel smart connect

Nonetheless, we will both know for sure after I post the screen shots on here.

Puppies04,

I will triple check the HDD cables and if the problem persists after doing what chusteczka wrote I will do HDD diagnostics. Maybe even do the MoBo reset as suggested.

I once again would like to say thank you both, you have made my day!

-Bruno

 

[brun0]

P.S. Puppies04, I was happy to read your reply as well! Haha

 

[chusteczka]

Write your bios settings down on paper. Then create an Excel (or OpenOffice) spreadsheet of your settings. Then print that out on paper and use it to write down each bios setting change as you go. This helps to remember where you have been and what you are testing from. Otherwise, tweaking bios settings becomes more of a gamble.

Unless you are able to change your bios settings through a software utility. Although a spreadsheet may still help.

Good luck.

 

[brun0]

The following link has all screenshot pages of my BIOS:
https://docs.google.com/folder/d/0B64F0R_gdveXb0tzeVlPaDBOSkE/edit

I will be posting in regular list format as soon as I finish writing this post, but just in case you wanted to see what I am seeing, you now have the visuals

 

[brun0]

Took a while, but I tried to be as thorough and detailed as possible:

AI Tweaker

MAIN:

1. Ai Overclock Tuner → Manual
2. BCLK/PEG Frequency → 100.0
3. ASUS MultiCore Enhancement → Enabled
4. Turbo Ratio → By All Cores
5. By All Cores (Can Adjust in OS) → 42
6. Internal PPL Overvoltage → Disabled
7. Memory Frequency → DDR3 - 1600 MHz
8. EPU Power Saving Mode → Disabled
9. OC Tuner → OK
   DRAM Timing Control

1. DRAM CAS# Latency → 9
2. DRAM RAS# to CAS# Delay → 9
3. DRAM RAS# PRE Time → 9
4. DRAM RAS# ACT Time → 24
5. DRAM COMMAND Mode → 2

CPU Power Management

1. CPU Ratio → Auto
2. Enhanced Intel SpeedStep Technology → Disabled
3. Turbo Mode → Enabled
4. Turbo Mode Parameters → Auto (ALL OF THEM)**

DIGI + VRM

1. CPU Load-Line Calibration → Ultra High
2. CPU Fixed Frequency → 350
3. CPU Power Phase Control → Extreme
4. CPU Power Duty Control → Extreme
5. CPU Current Capability → 120% (Max Amount)**

Main (Cont'd):

1. CPU Voltage → Offset Mode
   1. CPU Offset Mode Sign → +
   2. CPU Offset Voltage → 0.040
2. DRAM Voltage → 1.500V
3. VCCSA Voltage → Auto
4. PCH Voltage → Auto
5. CPU PLL Voltage → Auto
6. CPU Spread Spectrum → Disabled

Advanced

CPU Configuration:

1. Intel Adaptive Thermal Monitor → Enabled
2. Active Processor Cores → All
3. Limit CPUID Maximum → Disabled
4. Execute Disable Bit → Enabled
5. Intel Virtualization Technology → Disabled
6. Hardware Prefetcher → Enabled
7. Adjacent Cache Line Prefetch → Enabled
8. CPU Power Management Configuration
   1. CPU Ratio → Auto
   2. Enhanced Intel SpeedStep Technology → Disabled
   3. Turbo Mode → Enabled
   4. CPU C1E → Enabled
   5. CPU C3 Report → Disabled
   6. CPU C6 Report → Disabled
   7. Package C State Support → Auto

PCH Configuration:

1. High Precision Timer → Enabled
2. Intel Rapid Start Technology → Disabled
3. Intel Smart Connect Technology → ISCT Configuration → Disabled

APM:

1. Restore AC Power Loss → Power Off
2. Power on By PS/2 Keyboard → Disabled
3. Power on by PCIE/PCI → Disabled
4. Power on by Ring → Disabled
5. Power On by RTC → Disabled

System Agent Configuration:

1. Memory Remap Feature → Enabled
2. Graphics Configuration

(iGPU Frequency → N/A)

1. 1. Primary Display → PCIE
   1. iGPU Memory → Auto
   2. Render Standby → Enabled
   3. iGPU Multi-Monitor → Disabled
2. NB PCIe Configuration

(PCIEX16_1 x16)

1. 1. PC1EX16_1 Link Speed → Auto

(PCIEX16_2 Not Present)
2. PCIEX16_2 Link Speed → Auto

SATA Configuration:

1. SATA Mode Selection → AHCI
2. S.M.A.R.T. Status Check → Enabled
3. Hot Plugs → Disabled (ALL)**

Onboard Devices Configuration:

1. HD Audio Controller → Enabled
2. Front Panel Type → HD
3. SPDIF Out Type → HDMI
4. PCI Express X16_3 slot (Black) Bandwith → X4 Mode
5. Realtek LAN Controller → Enabled
6. Realtek PXE OPROM → Disabled
7. Asmedia USB 3.0 Controller → Disabled

Network Stack:
1. Network Stack → Disabled

 

[brun0]

Should I now be able to go back up to 1080i? It isn't my 19'' HDTV's "recommended" resolution according to windows 7 when I select resolutions, but I know my HDTV is capable of running it at 60Hz Rate.

Thanks again,
Bruno

 

[chusteczka]

Disclaimer: I am not familiar with Intel systems. If anyone else can help here, please step in.


PCH Voltage
Intel Z77 'Panther Point' Chipset Overview

Today's chipsets are merely responsible for I/O and have little bearing on performance. In fact, they're not even really a "chipset" anymore as there isn't a set of chips. Years ago, it was common for chipsets to have a northbridge and southbridge. The former was the direct line of communication between the CPU, GPU and memory, while the latter handled everything else from expansion slots to storage.

However, as features such as the memory controller were moved from the northbridge to the CPU itself, the chip played less of a role. Eventually, Intel took what was left of the northbridge and added it to the southbridge in what the company calls a PCH or Platform Controller Hub.

Therefore, we should not need to adjust the PCH voltage.


CPU PLL Voltage
What does PLL Voltage do?

Apparently, PLL voltage is an electrical engineering term for maintaining synchronization of the clock cycles between the originating component (the processor) and the reference components (i.e. front side bus (FSB), PCI-E, southbridge (PCH), ...). It is best to leave this on Auto. They mention the maximum safe value to be 1.7 volts.


VCCSA Voltage
http://www.tomshardware.com/forum/266192-29-vccio-vccsa

user: ortoklaz
VCCSA: More commonly known as System Agent Voltage. Intel’s maximum System Agent Voltage (Vccsa) is 0.971v and minimum is 0.879v. Stock is 0.925v. System Agent Voltage should NOT be touched, it is supposed to be a fixed voltage, and it powers many things that the VCC does not power. One of the most important is the Power Control Unit (PCU) which controls internal power allocation of the processor. This voltage is to be generated by a separate VRM than used for SVID.

It seems best to leave the VCCSA Voltage at Auto also.

=================

Read the reviews I provided above and look for differences between your setup and what they provide. I was not able to find a voltage setting for the PCI-E graphics bus.

Intel Z77 Motherboard Review with Ivy Bridge - ASRock, ASUS, Gigabyte, and MSI
ASUS P8Z77-V Pro – BIOS
ASUS P8Z77-V Pro - In The Box


google.com - asus z77 overclocking guide

overclock.net - ASUS Z77 Motherboards UEFI OC guides

ASUS Z77 UEFI Tuning Guide for overclocking:
http://www.mediafire.com/file/l34d6j6c4wqfx69/Z77_UEFI_OC_Tuning_Guide.pdf


Maximus 5 Gene Overclocking Guide/Tips:
http://www.mediafire.com/file/1p63cctaoap6gtt/M5Gguide.zip

rog.asus.com - ASUS Z77 Motherboard UEFI BIOS Tuning Guide

benchmarkreviews.com - ASUS UEFI Steps


It seems the information is out there to help with overclocking. What you want to do is double-check your settings with their recommendations. Also, look for something that controls the voltage to the PCI-E bus or the graphics.

I am personally not familiar enough with Intel to help any further.

If you are unable to find significant difference between your bios settings and the recommended settings, then try Puppies04's suggestions.

 

[brun0]

Thanks again,

I will look over all of the information when I get home from College tonight.

I will let you know of any changes/ questions that arise from the information.

I was finally able to register over at Sapphire Forums (It kept giving me errors yesterday). I have posted the thread in the following link if you want to follow that also:

http://www.sapphireforum.com/showthread.php?29780-Problems-with-quot-random-quot-gaming-crashes

-Bruno

 

[brun0]

I played the game for over an hour and it did not crash. The only change I did was lower CPU overclock to 4.2 ... but there are times i can play for hours and it crashes then, so I will play again tomorrow and see what happens.

Just wanted to fill you in

 

[brun0]

Played the game for maybe 2 or 3 hours, it crashed. The CPU was clocked to 4.2, only change made from before and I guess not the sole problem.
The following is a link to the core temps (using core temp):
https://docs.google.com/spreadsheet/ccc?key=0Aq4F0R_gdveXdEU0WEYtWUN0aEV3R25oQ3JqMFp1TEE

The following is the GPU-Z sensor log:
https://docs.google.com/document/d/11f4W_J-ih5juiytCS5Sxx0lplh-YZP-W6ap6DaVGGfQ/edit

I tried to log HWMonitor but didnt work out, I will re-log all three (CoreTemp, GPU-Z and HWMonitor Pro) and repost everything including al of the fancy graphs from HWMonitor

 

[brun0]

I have already sent a support ticket to Sapphire to send me the latest 7870 drivers as suggested over in the sapphire forums.. I will also be sending a support ticket to Asus. Will let you know what their replies are

 

[chusteczka]

Have you tried returning the bios to its default settings? This should create a stable system from which you can then see if the crashes continue. This is what I would do before creating a support ticket.

 

[BrightCandle]

Default driver settings, stock clock speed on everything , latest drivers for every component and latest bios. Once you have that done then you can test for a possible hardware fault.

 

[Puppies04]

Do not try to fault find while overclocked

I can't stress this enough, if you don't run at stock settings it is going to make it incredibly hard to find out what your problem is.

 

[Tintan]

In this upgrade your bios and graphic driver and good to go.

 

[brun0]

Hello all,

Regarding default settings: ok, will do

Tintan, what did you mean by your post?

 

[chusteczka]

Hello all,

Regarding default settings: ok, will do

Tintan, what did you mean by your post?

He meant to ensure you are using the latest bios and the latest graphics drivers.

Asus Support shows the lates bios is v0812 for your Asus P8Z77-V LK motherboard.
http://support.asus.com/Download.aspx?SLanguage=en&m=P8Z77-V LK&p=1&s=39


Your bios picture shows you are using the latest bios v0812.
https://docs.google.com/folder/d/0B...edit?pli=1&docId=0B64F0R_gdveXQmVoY1l6UnpzUUU


Your video card:
SAPPHIRE Radeon HD 7870 GHz Edition OC 2GB 256-bit GDDR5 PCI Express 3.0 x16 HDCP Ready CrossFireX Support Video Card (11199-03-20G)

Sapphire Drivers
http://www.sapphiretech.com/presentation/downloads/?pid=1518&psn=0006&lid=1&os=6

Just make sure you are using the latest set of Sapphire drivers from August 16, 2012.


Then reset the drivers, hardware, and bios to their default settings. Then test.