安裝 Steam
登入
|
語言
簡體中文
日本語（日文）
한국어（韓文）
ไทย（泰文）
Български（保加利亞文）
Čeština（捷克文）
Dansk（丹麥文）
Deutsch（德文）
English（英文）
Español - España（西班牙文 - 西班牙）
Español - Latinoamérica（西班牙文 - 拉丁美洲）
Ελληνικά（希臘文）
Français（法文）
Italiano（義大利文）
Bahasa Indonesia（印尼語）
Magyar（匈牙利文）
Nederlands（荷蘭文）
Norsk（挪威文）
Polski（波蘭文）
Português（葡萄牙文 - 葡萄牙）
Português - Brasil（葡萄牙文 - 巴西）
Română（羅馬尼亞文）
Русский（俄文）
Suomi（芬蘭文）
Svenska（瑞典文）
Türkçe（土耳其文）
tiếng Việt（越南文）
Українська（烏克蘭文）
回報翻譯問題
討論區規矩與守則
檢舉此篇文章
My GTX 780 runs all games at 1440p on high/ultra settings without lag, averaging 60fps+ and manages a temperature of 35 degrees. Using GeForce Experience with the latest GeForce Drivers. LED Visualizer is turned off on it, due to the fact you can't see it, the case is surrounded in noise foam and enclosed. You will probably find those artifacts are from overheating? Clean out the dust and give more airflow, then stop with the unnecessary overclocking? Overclocking is so 1999.
In fact, don't post if you're not even going to read the original post all the way through.It seems like you read the title and "overclock" and just stopped reading so you could reply.
On top of that you instantly assume my case is full of dust?
Yes, since you mentioned artifacts occured after your overclocking - I assumed 1+1=2
All I was trying to say is if it's artifacts due to prolonged overheating, and it doesn't go away, then your best bet is to return your card to the store under warranty (making no mention that you already broke that warrenty due to overclocking which is outside it's law).
No, I mentioned that I had artifacts even on stock factory clocks.
And again, reverting to previous drivers seemed to fix it. Please read before you type.
By the way, you don't seem like you know anything about safe temps. The GTX 780 automatically overclocks itself to achieve 80c. The fact that mine never goes above 65c under full load means it has really good airflow and cooling. By what you told me about your GTX 780, you don't actually have one. My brother has the same card as me, hasn't overclocked and is running on a 1440p monitor, his FPS is around the 40FPS range on ultra in BF4. And 35c is idle temps. Unless you're water cooling, I doubt your temp is anywhere below 60c.
Overheating is the most common cause of artifacts and the critical first point to check and ensure is fixed before moving on with any more testing.
So as you might of been offended and thought I didn't have any idea. You did ask and I was merely following safety first trouble-shooting for you.
The next check would be memory - both system RAM and graphic card VRAM.
Then PSU juice, if being drained by other devices and not having enough for fully powering the graphic card.
Lastly check software.
In most case, artifacting can also be a sign of a dying microchip, over time if not resolved will have your entire display blank out.
Being covered by warranty is an extremely good option at this point if it was a hardware issue. They would of replaced it for you for free. However, if they find out it's been overclocked, that voids all warranty.
Let me elaborate. Before overclocking or doing anything to it through software, it would run at 72c full load stable. Stock EVERYTHING It wasn't climbing.
I never said that my card was running at 72c while overclocked. In fact I've never seen it go over 68c WHILE overclocked. I did say it had been run at 72c (yes i am going to nitpick here.) and that it was before any overclocking or tinkering had ocurred.
Also, this has pretty much been solved by a simple driver downgrade so you telling me even more fixes and explainations for my artifacting is pretty much redundant and a waste of space on valves forum servers.
If you are OC'ing and u get artifacting; something within the settings are wrong.
Either a Clock setting is too high; the Voltage is too low; or the temps are too high with the OC and thus u may need to manually raise the GPU fan % to compensate.
Now if artifacting occurs on a GPU running at default clocks/settings; return it! As most likely it has a faulty GPU Core or VRAM.
Use software such as EVGA Precision or MSI Afterburner to OC with. If you notice such issues; quickly close what you're running and Reset the clocks back to default. Do not let the app continue to run, or the OC'ed GPU clocks/voltage to remain applied when u see visual artifacts; you could end up damaging the GPU over time.
I downgraded to 335.28 and my artifacting instantly stopped. Stock and OC
As sometimes (and its been reported by many owners) even changing from one NVIDIA GPU to another; it may help to clean install the NVIDIA Drivers again (even the same version). So this could explain your "solution". So I would perhaps try uninstalling the 335 drivers and clean install the 337.88 WHQL and see if that helps too.
To clean install an NVIDIA GPU Driver; simply choose Custom during the installer screen and tick the "Perform Clean Install" box.
I would also go on EVGA.com and Register your Free Account (if u haven't already) and register your 780 for warranty purposes. If such issues arise in the future, I would take that up with EVGA and see what they recommend. Also be sure to look through the EVGA Forums > http://forums.evga.com/EVGA-GeForce-700TITAN-Series-f87.aspx
I have already done a clean install of 337.88WHQL and those are the drivers that gave me problems. As I stated before, I did a clean downgrade to 335.23 and haven't had any issues since.
Another friend of mine with a GTX 660 had some graphical bugs on some games we play together on 337.88 and he downgraded to 335.23 and didn't have any issues anymore.
It seems like Nvidia gave us the short end of the stick when it came to 337.88 WHQL drivers optimized for Watch_Dogs. (Supposedly)
EDIT: I've just registered my GTX 780 on EVGA's website. I'll be sure to keep you informed if anything happens. As of now I'm only having performance issues with PhysX, but from what I've been told, everyone does.
Manufacturer NVIDIA
Model GeForce GTX 780 Ti
Device ID 10DE-100A
Revision A2
Subvendor ASUStek Computer Inc (1043)
Current Performance Level Level 1
Current GPU Clock 324 MHz
Current Memory Clock 324 MHz
Voltage 0.887 V
Bus Interface PCI Express x16
Temperature 25 °C
Driver version 9.18.13.3788
BIOS Version 80.80.34.00.10
Physical Memory 3071 MB
Virtual Memory 3072 MB
I've tried out a couple we have laying around here, including my own personal GPUs that I use and haven't seen such issues with 337.88 WHQL. And I and some co-workers are not only using this driver, but have ran Unigine Heaven and Valley (as we always do with each GPU test) and we also play Watch_Dogs. No real issues to speak of with the GPUs listed below.
GPUs such as:
EVGA (SC versions) GTX 550-Ti, 570, 660, 770, 780 Ti
MSI (Gaming Editions) GTX 750-Ti, 760 2GB (shorter ITX version), 770 4GB
Course we play alot of other games too, new and old, so take it with a grain of salt...
Not to say what u saw or experienced was a fluke, but it happens. Hopefully in your case with your GPU, the next NVIDIA driver version will be more stable for you. 334/335 Drivers are still fairly good, so just try the next WHQL Version when it comes and see how that does.
So @mAGNUSwAKER - No you won't get this issue, in fact it's really rare to occur.
Bad-Motha has a good point about the older AMD drivers still being around, but if you clean installed the OS that's not the problem.
Graphic card artifacts:
#1 reason is overheating issues caused by overclockers
(not saying it's the only reason, but 90% of all cases)
I've seen similar artifacts on other people's machines in the past when they have been attempting to overclock at the absolute verge of stability and pushed just a bit too far.
What had occurred is even when the card itself had dropped back down in temperature, reported by the monitor, the video card memory VRAM was still actually extremely hot.
When overclocking the memory as some of the cards memory chips may not be making good contact with the thermal pads under the heat sink, meaning one or two memory chips may be getting much hotter than the others. This doesn't get monitored very well.
You can use this OC Scanner for detection of artifacts:
http://www.evga.com/ocscanner/
Over time if it goes away, then that could be fine. However if it stick around you have physical damage done. For the graphic card video memory - It appears just like that video you showed. Sometimes underclocking can help. But most of the time bad video RAM is permanently damaged as a small area on one or more of the chips, and you just have to replace the video card.
It doesn't have to be your video memory either, it could be system RAM, but with that crashing also occurs, etc.
If it was the GPU, then the artifacting goes crazy to the point that nothing on the screen is recognizable. So it's not that, more likely memory.
EVGA claims that overclocking of their graphic cards doesn't break warranty... with the extra loop hole cause, so long there's 'no physical damage done' to the card. Therefore extreme/crazy overclocking isn't covered if it fried a chipset or your memory. If it does continue and you have to get a replacement from warranty - you would have to claim that it was like that when you purchased it, make no mention about overclocking and hope for the best.
@Fespie - While this might of not been your case, I personally believe it is and fingers crossed it was temperature and not damaging. I say it only as a warning, maybe not just for you but also others doing the same. You can still choose to ignore it's warning. I'm posting only for other people to advice. Overclocking isn't that bad if doing small step increments as you can quickly revert issues like this without perm damage. Probably not a waste of space on valves forum servers, rather just a waste of space for your 'murica paranoid NSA spy to collect. So sue me for trying to help.