Sign in with
Sign up | Sign in
Your question

Unigine Heaven 4 black screen

Tags:
  • Intel i5
  • Graphics
Last response: in Graphics & Displays
Share
September 27, 2014 12:57:39 AM

Around scene 6 to 7, a black screen where my monitor displays no signal pops up. At this point it will either crash or pop back up, stutter and continue. One thing to note is I consistently get a read of 4.5 fps as my minimum read from this occurrence.
It happens in several other scenes past it, but I haven't kept track because I get frustrated watching a choppy benchmark.
It happens regardless of SLI enabled or disabled.

Heaven used to run just fine 6 months back. Any advice?

Specs:
i5 3570k @ 4.1ghz
asrock extreme4
16gb ddr3
2x 660ti
750w Coolermaster
120gb vertex 4 (ssd boot drive)
windows 7 home premium 64 bit

More about : unigine heaven black screen

September 27, 2014 1:11:45 AM

Looks like the PSU is passing out, CM is not a great PSU manufacturer, and 660 Ti in SLI is really putting high load on it, which is exactly why wise people don't get CM or FSP or Logisys or Raidmax or Corsair CX/ CS/ VS for high loads, they use cheap capacitors which melt on high load. That's what I suspect is happening with your system. Getting a good 750W PSU like EVGA 750W B2 for $50 from Newegg would solve the issue.
m
0
l
September 27, 2014 1:27:59 AM

MeteorsRaining said:
Looks like the PSU is passing out, CM is not a great PSU manufacturer, and 660 Ti in SLI is really putting high load on it, which is exactly why wise people don't get CM or FSP or Logisys or Raidmax or Corsair CX/ CS/ VS for high loads, they use cheap capacitors which melt on high load. That's what I suspect is happening with your system. Getting a good 750W PSU like EVGA 750W B2 for $50 from Newegg would solve the issue.


You do sound confident about the power supply being the issue, but I'm not too sure. In running games, such as far cry 3. I can run fine for a prolonged amount of time even in SLI.
Crysis 3 crashes a bit though, but I need to do more experimenting to see if it's software related, I got a cheap Russian version. It ran fine with no hiccups once again back 6 months ago.

I want to hear more opinions.
m
0
l
Related resources
September 27, 2014 1:39:51 AM

Crysis 3 is more GPU intensive than FC3, and I'm pretty certain it'd crash on FutureMark GPU testing too.
m
0
l
September 27, 2014 3:35:39 AM

MeteorsRaining said:
Crysis 3 is more GPU intensive than FC3, and I'm pretty certain it'd crash on FutureMark GPU testing too.


I tried running 3DMark11, it took a while, but I got the results. Both single card and SLI ran just fine, no hiccups.
Although I'm using the demo, I still think the results are relevant.

I'm beginning to think it could be a software issue, rather than hardware. I did an online watt calculator and I should be using 600w at most. I understand coolermaster isn't the best psu brand, but it isn't mediocre either.

I'm running Nvidia drivers 344.11, but having tried several other drivers, heaven is still not having it. Strange...
Any more tests to recommend?
m
0
l
September 27, 2014 3:58:34 AM

I'm sorry to say, but unless you have V or VS or GX PSU, it indeed is below average brand in PSUs. 600W is actually high for 750W CM. You see, the efficiency as we read from PSU stickers is not for decoration, and 80% efficiency which is of most CM PSUs, for a 750W means it can give 600W at max. So going by that, it'd not be a huge surprise if it turns out to be the PSU.
Also, try running Prime95 while running 3DMark11, it'll crash, certainly, if it doesn't, I'd be satisfied its not the PSU.
m
0
l
September 27, 2014 3:58:36 AM

I have a feeling one of your cards is slightly faulty, and you're only seeing some problems when it's under very high load.
What coolermaster PSU do you have?
m
0
l
September 27, 2014 4:29:50 AM

It is a coolermaster GX 750w, sorry I didn't specify. It's bronze certified 80% efficiency.

When I said 750w, I thought decent model was assumed.
m
0
l
September 27, 2014 4:47:24 AM

GX isn't decent unfortunately :(  It'd be good to upgrade, but I'm not sure it's the cause of your issues.
m
0
l
September 27, 2014 5:15:38 AM

Okay, I don't know if it's just you guys, but I feel that you have high standards in PSUs. I even have a 450w 70% efficiency from thermaltake that still runs solid up until today.

I know it's not a Corsair or Seasonic, but I don't think it's significantly worse.
It's the brand that's docking me points on the PSU right? Cause the specs look good, SLI certified, 85% typical, 12V rail 60A, and 5 year warranty.

Anyways I ran 3DMark again, SLI and I done it several times. It ran rock solid. If it was a PSU issue wouldn't all programs be crashing?
I wasn't taking it easy on the cards either, ran 100% on both for 2 hours. I'm suspecting software related.
Other things I suspect could help.
Reseating my cards, ensure they are snug. Check all connections. Maybe I got a bad SLI bridge? Maybe it's not in firm enough..

Oh and yes I ran prime95 at the same time!
m
0
l
September 27, 2014 5:47:13 AM

I don't think it's a PSU issue, but I'm just letting you know that it's not well rated at all.
It could well be the SLI bridge, is it a flexible one? I still think it may be an issue with one of the cards but we should rule everything else out first.
m
0
l
September 27, 2014 6:06:43 AM

Well, some additional information on the cards.

I have a GTX 660ti Gigabyte Windforce bought new January 2013. And up until this day, this card ran just fine. The temp max were the same as it was back when I bought it!

2nd one I bought last week an GTX 660ti Asus DirectCU II, used from eBay. It looked good, the temperatures were good. I never used this card outside of SLI. The previous owner said it was in perfect condition, good as new and I believe it.

I could try to test cards alone and see if that has to do with anything. This black screen thing only occurred since I installed my second card, but to be fair I haven't tried heaven in 4 months at least, so I don't want to jump to conclusions.
m
0
l
September 27, 2014 6:08:09 AM

Hmmm, a lot of the time when people sell 'perfect' cards on ebay they have problems like artifacting and black screening. Just being honest, I hope that's not whats happened.
m
0
l
September 27, 2014 6:13:08 AM

He was a reputable seller, 100% positive with 71 transactions all dealing with computer parts. Also the black screen issue as far as I know is quite specific to Heaven. Not sure if I should call it black screen, but you know when your monitor states "no signal" for a second. That's what I'm seeing.

If there is a step by step troubleshooting guide for me to follow, I'm all ears.
m
0
l
September 27, 2014 9:11:26 PM

Sorry for late reply, different time zones you know. Well, its not the PSU, but certainly the GPU. The descriptions you provided point it out there, is it still under warranty?
m
0
l
September 27, 2014 9:13:48 PM

IDK about 4.0 , but 3.0 would not display on my 1440p monitor if it was set at native, i had to select 1080p
m
0
l
September 27, 2014 9:54:06 PM

MeteorsRaining said:
Sorry for late reply, different time zones you know. Well, its not the PSU, but certainly the GPU. The descriptions you provided point it out there, is it still under warranty?


Both are within their 3 year warranty.
m
0
l
September 27, 2014 9:56:58 PM

Replace the one you bought from eBay, it appears to be faulty, to me.
m
0
l
September 27, 2014 11:10:35 PM

MeteorsRaining said:
Replace the one you bought from eBay, it appears to be faulty, to me.


Too hasty, it works on other games and applications just fine. I want to be sure your right so I think I'll test each card on heaven individually.
Cause it'll be a big hassle to go through with a replacement and find out the issue wasn't hardware.
m
0
l
September 28, 2014 1:39:43 AM

Sure you should, I was assuming you've checked it already, nevermind.
m
0
l
September 28, 2014 10:00:11 AM

Okay an update. I removed my cards physically and tested my cards one by one. My Gigabyte to my surprise still did that no signal frame that drops to 4.5. The Asus had a worse effect. My rig kept restarting like it was rejecting a body part. Note I used the top PCI 3.0 x16 slot where my Gigabyte originally resides for both tests.
m
0
l
September 28, 2014 10:43:43 AM

False alarm with the Asus reseating fixed it
m
0
l
September 28, 2014 11:02:11 AM

So both have settled well now?
m
0
l
September 28, 2014 11:31:40 AM

Update with good news.

The Asus is not a faulty card, it works great, tried Heaven 4, the arguably the most finicky stress test on both slots, works fine.

The Gigabyte consistently gave me that black screen, so I tried what I hated..underclocking. The no signal/crashing stopped by -50mhz on the core... It's factory overclocked and I hate to underclock!

Next up I'll try SLI again :)  but I have a good feeling
m
0
l
September 28, 2014 12:22:54 PM

I think I got the root cause guys!

So I did an SLI run in Unigine Heaven 4.0 and just as I predicted the results were what I expected.

At default clocks, the black screen would happen and even artifacting, a clear sign something is clocked too high. The suspect being my gigabyte card.
I manually underclocked it in Afterburner by -50mhz on the core and did another SLI run. It ran solid with no "errors" as far as I can see.

What doesn't make sense to me is my cards are running SLI asynchonously. I thought SLI downclocked the faster clock to the slower one, my Asus runs 65mhz slower on the core than my Gigabyte.

In any case I think I can safely conclude this from our conversions:

-The PSU isn't an issue
-The SLI bridge is fine
-The card from eBay, even though it's used and I had it for a week, seems solid.
-Nothing wrong with my PCI 3.0 x16 slots
-Not an Nvidia driver issue
-Not an SLI issue, in fact SLI scaled good! About doubled my score as opposed to one.
-The Gigabyte card is clocked too high even though it's factory overclocked!

If you guys have any more integrity tests, I'm all ears. I care more about stability than raw frame rates.
Also is there a way for me to change the default clock on my Gigabyte so I don't have to underclock it everytime I restart my PC?
m
0
l
September 28, 2014 8:37:22 PM

I'm glad you could identify the root cause, all eyes were on poor Asus! GB seems to have done a bad job checking for stability, but did you have the issue before you got the Asus card? That confuses me. Anyways, I'm not Nvidia user currently but if I'm not wrong, Afterburner does have some sort of permanent settings, not sure where, but they must be in clock speed or something similar, like a check box.
m
0
l
September 29, 2014 12:15:04 AM

Yup, apply overclocking at system startup is the check box, hopefully it works like I expect. I'm calling Gigabyte tomorrow to see if there is any sort of compensation.
The PSU and used graphics cards were easy targets, but I refrain from kneejerk, prejudiced responses.
I do tests before concluding things. Maybe I should have provided further detail on the specs and condition of my parts to avoid misinterpretation next time.
It would have helped if you asked more questions, so I can know what you don't know. lol that sounded funny.
m
0
l
September 29, 2014 2:28:01 AM

Ohk, seems like we're almost done. Well, can't stress the point enough which is always on my sig!
m
0
l
!