Crysis just crashed on 1920x1200, Low settings, 0x AA.
I had one Corsair TR3X6G1600C8D Dominator module installed in the DDR3_1 Channel 0 input. (9-9-9-24-2, 1066MHz, 1.64V)
Even if High settings were asking too much from my system, crashing on Low settings is simply unacceptable.
YouTube - Crysis crashing on low settings
I had to physically turn off the system. Upon reboot, there were display artifacts in the BIOS. The system ultimately crashed attempting to load Windows 7 (the screen looked the exact same as my photos four posts back)
I had one Corsair TR3X6G1600C8D Dominator module installed in the DDR3_1 Channel 0 input. (9-9-9-24-2, 1066MHz, 1.64V)
Even if High settings were asking too much from my system, crashing on Low settings is simply unacceptable.
YouTube - Crysis crashing on low settings
I had to physically turn off the system. Upon reboot, there were display artifacts in the BIOS. The system ultimately crashed attempting to load Windows 7 (the screen looked the exact same as my photos four posts back)
Sounds like bad ram sticks to me, if changing ram settings had an effect on stability its the most likely option. i7 have onboard memory controllers right? If so then a very vague possibility its a poor chip. Motherboard/Chip combinations can be extremely picky about timing/voltage settings, so playing with them (both up and down) would be my first move. Are the programs you listed the only thing that you are using for stability testing? memtest86 (loaded via boot disk) is the standard for ram stability testing, but can take quite a while to expose problems (12hrs+), so could be useful. If it were me I'd use superpi 32M calcs to start assessing stability, with your current setup it will likely fail (memory intensive), so I'd mess with ram timings (try being conservative, 10-10-10-2X-2T with a low multi and work up, maybe even try dropping voltage as ram chips can be weirdly picky) and get 32M superpi (google it) stable. Then move to prime95 and select the memory intensive stress test. If the ram isnt solid this will fail rather quickly, but then go back and go further in the direction with each timing that you did to get superpi stable.
But in the end it could just be bad sticks... corsair have a great RMA system (I think they may even cross-ship, no pun intended).
I have to say I have not often experienced artifacting from unstable memory, so bear in mind you may have been extremely unlucky with your stock and got a bad graphics card too. I am aware that nvidia were quite behind stated release for 'fermi' the chip architecture of your 470, and most reviews found them to be power hungry, hot running cards, so they probably struggled with yield of good chips. This could make it a little more likely you got a bad sample.
But in the end it could just be bad sticks... corsair have a great RMA system (I think they may even cross-ship, no pun intended).
I have to say I have not often experienced artifacting from unstable memory, so bear in mind you may have been extremely unlucky with your stock and got a bad graphics card too. I am aware that nvidia were quite behind stated release for 'fermi' the chip architecture of your 470, and most reviews found them to be power hungry, hot running cards, so they probably struggled with yield of good chips. This could make it a little more likely you got a bad sample.
I'm not sure if the RAM is necessarily bad or not. The optimized default timing in the BIOS is 8-8-8-20-2 and 1.5V. My modules are rated for 8-8-8-24-2 and 1.65V.
The tRAS (cycle time) in the BIOS was much lower than Corsair's specs. I'm hoping that was the main problem. I've been running Prime95 and haven't gotten any errors yet.
Of consideration may be my +12V rail. I had Speedfan running during Prime95 testing and it showed my +12V rail bouncing between +12.5V and +13V with large voltage spikes. The voltage occasionally dipped below 11.5V and occasionally peaked above 13.5V.
The tRAS (cycle time) in the BIOS was much lower than Corsair's specs. I'm hoping that was the main problem. I've been running Prime95 and haven't gotten any errors yet.
Of consideration may be my +12V rail. I had Speedfan running during Prime95 testing and it showed my +12V rail bouncing between +12.5V and +13V with large voltage spikes. The voltage occasionally dipped below 11.5V and occasionally peaked above 13.5V.
Hi Thadman, I think you will find that the new card fixes the majority of your issues, but the ram timing is another thing that could be causing problems. Do the ram modules have spd (not sure if the latest memory does) and have you a setting in the bios to use spd function? If they do then it should ensure that they run at their stable settings. I'm a bit out of date on the latest ram technologies (don't know what all those timings are) but I would think if the ram is rated 24 and you run it at 20 you are running it with some parameter faster than what it is reliably capable of, and doing so could result in instability. I doubt the display corruption is related, but certainly hangs or crashes could be.
Tony.
Tony.
For some reason, CPU-Z thinks my Corsair Dominator TR3X6G1600C8D modules are actually CM3X2G1600C8D modules...thats weird
false alarm. I contacted Corsair, CM3X2G1600C8D is the module number and TR3X6G1600C8D is the kit number.
I find it interesting how sporadic the instability of the system is. It will be totally unusable for a period...and then all of a sudden work flawlessly. Then it will get progressively worse and then ultimately become unusable again. Then all of a sudden it will work flawlessly...and the cycle will continue. I was able to access the desktop for about 2 hours today until it crashed...and then continued crashing until I was unable to boot into the desktop.
I'm now totally unable to boot into the desktop if I have GFX drivers installed. As soon as the OS begins to load, my GPU fan kicks into overdrive and the system freezes. I'm only able to boot into safe mode. This is irrespective of the timing settings of my RAM modules in the BIOS (8-8-8-20-2 and 8-8-8-24-2 produce the same results).
Whats interesting is that now my audio output is not working.
I clicked on the sound icon and troubleshoot sound problems and received this error "one or more audio service isn't running".
I have not uninstalled any motherboard drivers recently and the audio was previously working...
These past two weeks have been very frustrating. However, I will say, I am learning a LOT about diagnosing problems (software as well as hardware) with the help I've received in this thread. The next time I experience a computer problem, I'll be much more prepared.
Also, thanks for all the replies everybody. I really appreciate it.
It's a lot easier to do all this troubleshooting with advice from others who have experience with similar problems.
I'm now totally unable to boot into the desktop if I have GFX drivers installed. As soon as the OS begins to load, my GPU fan kicks into overdrive and the system freezes. I'm only able to boot into safe mode. This is irrespective of the timing settings of my RAM modules in the BIOS (8-8-8-20-2 and 8-8-8-24-2 produce the same results).
Whats interesting is that now my audio output is not working.
I clicked on the sound icon and troubleshoot sound problems and received this error "one or more audio service isn't running".
I have not uninstalled any motherboard drivers recently and the audio was previously working...
These past two weeks have been very frustrating. However, I will say, I am learning a LOT about diagnosing problems (software as well as hardware) with the help I've received in this thread. The next time I experience a computer problem, I'll be much more prepared.
Also, thanks for all the replies everybody. I really appreciate it.
It's a lot easier to do all this troubleshooting with advice from others who have experience with similar problems.
Last edited:
These are indeed weird problems.
Did you run memtest to check your memory? I think you did not mention it, so I thought I might ask. You can download it here Memtest86+ - Advanced Memory Diagnostic Tool.
There is a bootable DOS-ISO, so you can be sure that there is no windows driver interference.
Run it for at least 6 hours (longer is better), if it reports errors you can run it one stick at a time to find the faulty one
hope that helps
daniel
Did you run memtest to check your memory? I think you did not mention it, so I thought I might ask. You can download it here Memtest86+ - Advanced Memory Diagnostic Tool.
There is a bootable DOS-ISO, so you can be sure that there is no windows driver interference.
Run it for at least 6 hours (longer is better), if it reports errors you can run it one stick at a time to find the faulty one
hope that helps
daniel
I don't have a USB stick yet and the only burnable media I have access to are DVDRs (However, I don't have a DVD burner). I'll drive over to BestBuy, pick up a USB stick, and see if memtest produces any results.
What about power? Could it be that my PSU isn't able to source enough power from the wall to power my GPU? I live in a small campus apartment.
This video on [H]ardOCP.com shows the GTX470 pulling over 430 watts at full load (no OC).
Power - Temp - Sound - NVIDIA Fermi - GTX 470 - GTX 480 - GTX 480 SLI Review | [H]ard|OCP
The bit-tech review of the i7 930 shows it loves power.
i7 930 @ stock / idle = 124 watts
i7 930 @ OC (4.3GHz) / idle = 176 watts
i7 930 @ stock / load = 216 watts
i7 930 @ OC (4.3GHz) / load = 396 watts
Review - Intel Core i7-930 CPU Review | bit-tech.net
I've got to be pulling over 700 watts without OC. Would that produce the aforementioned errors?
Hi Thadman, I think it is established that the video card has a problem. Wait till you get the replacement, put it in, and THEN see if you still have any problems. You could be spending all of this time worrying about something that may only be a side effect of a faulty video card
Certainly my machine was suffering from pretty much all the symptoms (freezes crashes, display corruption, spontaneous reboots and failures to boot) that you are, and the only thing I did was replace the video card and it has been fine ever since
Tony.
Certainly my machine was suffering from pretty much all the symptoms (freezes crashes, display corruption, spontaneous reboots and failures to boot) that you are, and the only thing I did was replace the video card and it has been fine ever since
Tony.
There is no way a tx950 is underpowered for your rig.
Usually if you would draw too much power and have a good PSU your computer should simply shut down. My guess is a faulty GPU as well, but since you have problems with yourmemory as well it could also be your Mobo or your memory.
The power consumption shown by hardocp and bit-tech are for the whole system at the wall.
In some forum there is a thread showing all the tests that measure the power consumption of the video card alone - of the top of my head I remember ht4u and xbitlabs:
Gigabyte GeForce GTX 480 and GTX 470 Graphics Cards: DirectX 11 Battlefield (page 6) - X-bit labs
HT4U.net Charts: Leistungsaufnahme Grafikkarten - Leistungsaufnahme Grafikkarte: Furmark (Seite 5) - Tests bei HardTecs4U
These show 209 and 236 Watts respectively. Bear in mind these are occt burn test/furmark numbers, both programs are considered by some a power virus, as they push graphics card to the edge way more than any real life application. I'd say they are a good way to test if the manufacturers did their homework and implemented proper throttling mechanisms to protect the card.
The same applies to cpus unfortunately I can't remember where I read the article about the consumption of CPUs alone.
Usually if you would draw too much power and have a good PSU your computer should simply shut down. My guess is a faulty GPU as well, but since you have problems with yourmemory as well it could also be your Mobo or your memory.
The power consumption shown by hardocp and bit-tech are for the whole system at the wall.
In some forum there is a thread showing all the tests that measure the power consumption of the video card alone - of the top of my head I remember ht4u and xbitlabs:
Gigabyte GeForce GTX 480 and GTX 470 Graphics Cards: DirectX 11 Battlefield (page 6) - X-bit labs
HT4U.net Charts: Leistungsaufnahme Grafikkarten - Leistungsaufnahme Grafikkarte: Furmark (Seite 5) - Tests bei HardTecs4U
These show 209 and 236 Watts respectively. Bear in mind these are occt burn test/furmark numbers, both programs are considered by some a power virus, as they push graphics card to the edge way more than any real life application. I'd say they are a good way to test if the manufacturers did their homework and implemented proper throttling mechanisms to protect the card.
The same applies to cpus unfortunately I can't remember where I read the article about the consumption of CPUs alone.
Last edited:
While the PSU may not be overwhelmed it could be defective. That being said, repeat after me: " I will never take voltage readings from software or the bios". These voltage readings are often HEAVILY off. I have had the bios read a fluctuating 9-10V on the 12V line while a calibrated DMM read 11.93V stable.
You need to use a DMM to read the voltages because they could be way off or they could be quite true.
You need to use a DMM to read the voltages because they could be way off or they could be quite true.
I am pretty sure you will find a new video card will fix your problem. usually with bad RAM the computer will just crash when that part of the memory is attempted to be used by the operating system or programs. same thing with a power droop, once too much power is drawn from the PSU or it is weak, something will jsut freeze up, rather than continuously "almost working"
a possible reason it might have worked once is because when you power cycled it the video card cooled down and whatever bad connection there is was good for a while. (thousands of SMD chip solder points, any of the ones on the memory or GPU > memory bus could be flakey.)
and last but not least. if the video card is doing something stupid, usually its the video card thats broken
a possible reason it might have worked once is because when you power cycled it the video card cooled down and whatever bad connection there is was good for a while. (thousands of SMD chip solder points, any of the ones on the memory or GPU > memory bus could be flakey.)
and last but not least. if the video card is doing something stupid, usually its the video card thats broken
I 100% agree. 95% or higher chance it is the video card, 5% or lower it is something else, such as PSU is faulty and the video card is power starved, etc. I just threw my info in because I can't stand to see people trust the voltage readings from software or bios.
I am pretty sure you will find a new video card will fix your problem. usually with bad RAM the computer will just crash when that part of the memory is attempted to be used by the operating system or programs. same thing with a power droop, once too much power is drawn from the PSU or it is weak, something will jsut freeze up, rather than continuously "almost working"
a possible reason it might have worked once is because when you power cycled it the video card cooled down and whatever bad connection there is was good for a while. (thousands of SMD chip solder points, any of the ones on the memory or GPU > memory bus could be flakey.)
and last but not least. if the video card is doing something stupid, usually its the video card thats broken
Something also to consider regarding your instability, a faulty SSD drive. My system is pretty similar to yours - the beefy video card, same case, psu, ssd. After a year my X25 ssd has started emitting a high pitched noise which now I am reading is due to faulty capacitors in the drives. A few seconds after the beep, system crashes, mostly when playing games or watching videos. I ended up migrating my system to boot from an external sata drive just to narrow it down. Anyway something to think about since you also use the same intel drive, some instability could be coming from that. I think everyone else is right about the video card, should try and use something else for testing.
Any of the artifacts are caused by some sort of faulty video ram. Hopefully you've started the RMA procedure.
The freezing is caused by not having the RAM at proper specs. Like you said before, you should operate them at 1.65volts @8-8-8-24 1600mhz or 1.65volts 9-9-9-24 at 1333Mhz.
The error "one or more audio service isn't running" is caused by being in safemode.
Then again i could be that by using a certain USB port that causes the artifacts.
Nah, I made a computer that had some serous problems when you used a certain USB port at the back.
The freezing is caused by not having the RAM at proper specs. Like you said before, you should operate them at 1.65volts @8-8-8-24 1600mhz or 1.65volts 9-9-9-24 at 1333Mhz.
The error "one or more audio service isn't running" is caused by being in safemode.
Then again i could be that by using a certain USB port that causes the artifacts.
Nah, I made a computer that had some serous problems when you used a certain USB port at the back.
without time to read the whole 12 pages, Try aiming a house fan into the case. If its a heat issue this should clear it to figure out what is heating up. , I am thinking from vid card, and a disgruntled chinese sweat shop worker didnt apply heatsink and compound properly.
The northbridge, and southbridge heatsinks getting hot, But once again, if they are hot to the touch right after booting its not good, however they usually just cause errors, and close out programs such as games and stuff. No artifacts that I have ever seen.
The northbridge, and southbridge heatsinks getting hot, But once again, if they are hot to the touch right after booting its not good, however they usually just cause errors, and close out programs such as games and stuff. No artifacts that I have ever seen.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.