Guild Wars 2

I run a lot of solo PvE and have been using Sinister on my necro with reaper runes. I need to spec some toughness where I typically run solo. Previously I was running Dire gear with Undead runes. Maximum CD, maximum survivability, essentially no effective DD to speak of…but with a staff a great build for defending towers in WvW. And in PvE, while it might take longer to bring down tough foes, with the Dire gear I can typically do it without worrying too much about dying.

That said, I’ve been wanting to move my necro/reaper to a build with more power, without giving up too much of the CD that makes a necro a necro, so chose the Sinister set. The way I see it, I MAY have gone with the Viper gear had I known about it, but it’s not enough of a gain over Sinister that it’s worth spending a pile of gold on. Perhaps as a long term goal, it might be worthwhile, but for the time being, nah.

Rauderi.8706:

Also, rising gem value freezes assets that were otherwise going to remain in game from irrepressible gold-farmers gaining loose cash from dungeons and other sources of liquid currency. All buying gems does is re-liquefy that for the buyer, who then injects that back into the economy, with a hefty service charge similar to the Black Lion Trading Post. Those fees remove huge chunks of gold at a time, first on the gold->gem transaction, and again on the gem->gold transaction.
All while giving ANet the real-money infusion they built the system for in the first place.

But what I’m saying is that there are far more gems being purchased with real money and converted to gold than there are gems being created from the gold being converted to gems and ANet will certainly not stop selling gems, so even after taxes, far more money comes in than goes out. Yes, there are other forms of liquid currency in the game, dungeons were very lucrative for a select few, but overall vendors probably create the most “new” in game gold, even more than the gem to gold conversion. It’s too bad GW2Spidey doesn’t give gem in/out volumes along with the historical trade value and ANet as long since removed the graphs from the interface. The only time I ever hear of anyone converting gold to gems is to pay for world transfers and even then, most people I know opt to use real money simply because the gold is worth too much now. It’s not like the early days, when you could buy 100 gems for 2 gold. Of course, in those early days few people were buying gems with real money because the gem to gold conversion was so bad that it simply wasn’t worthwhile. 100 gems costs over 20g at them moment. Few people, even the uber rich are going to pay that kind of money. I know I wouldn’t have the gold I have if I wasted it like that. However who does high prices like that benefit? High prices means big gold in game, means people are buying gems, means ANet wins, which isn’t a bad thing, except that it contributes to inflation.

Anyone who thinks people are converting enough gold to gems to cover the amount of gems purchased with real money is deluded.

Rauderi.8706:

Not that I understand the qqsalt over people buying gold with money (in a secure, legal way). They bring gem prices down so other people don’t have to grind as much to afford gem store items.

You are mistaken, buying gold with real money does not bring prices down. This issue with buying gold with real money is that it causes inflation in-game because it puts new money in to the economy that never existed. The more money available, the more expensive things get, especially uber rare big ticket items. Now before anyone gets on my case about people converting gold to gems and that’s were gems come from, so there is a limited amount. That may have been true, to a point, a couple years ago but do you think ANet (NCSoft) is going to say no to someone who want to give them some money? There will also be an artificial floor on what gems to gold will be worth simply because if the value drops too much the only people who will buy gems are those looking to buy something in the gem store…and that’s NOT where the big money is. There are FAR more people buying gems with real money than there people buying gems with gold, simply due to the fact that you need gold to get the good stuff…not gems. I know lots of people who buy gems monthly, like a subscription service. In contrast, I haven’t heard of anyone putting gold in to gems in a long time.

Ironically, the only gold one can “buy” that has a minimal affect on the economy, is the illegal stuff the gold sellers sell simply due to the fact that that gold was already part of the economy and so can be considered “old money”. But this is dirty money acquired through hacked accounts or other nefarious means and should never be supported.

So a reduction of 20%, my card doesn’t have a shader clock apparently, so, reduced the GPU clock from 967MHz to 774MHz and the left the memory clock as is at 1502MHz. Played 3 hours with settings cranked and frame limiter unlimited. Had one brief moment of stutter that lasted about 5 seconds. I saw no noticeable reduction in performance and no crashes. It’s only a single test but looking hopeful.

I have a question about the logs generated by GPU-Z. The behavior is the same with the original clock speed as well as the reduced clock speed…so I guess I’m not sure what I’m looking at. Mid-game, on a busy map, say the Tarir event, GPU usage is pinned typically at about 99%, Power Consumption [% TDP] floats between 90% and about 103%, while temperature sits pretty steady between 80-83C, and fan speed sits around 60%. Core clock sits around 914MHz under load as opposed to about 1120MHz with the original clock speed. Is this what I should be expecting to see?

abomally.2694:

If you’re interested, I use an older version of EVGA Precision (didn’t like the newer version) – I uploaded it here:

http://www.filedropper.com/evgaprecisionsetup212

It’s version 2.12. You can set it to apply your settings at start up. It applies the settings and then exits, so it isn’t running in the background. I’m using it on Windows 10 64 bit. Below is a picture of it.

Thanks, I’ll check it out. The one I got from eVGA for my card, 6.x I think, doesn’t like Windows 10 at all. It starts after throwing some driver handle error, then settings don’t get applied to the card when hitting apply. I was looking at Afterburner, but I don’t like how the settings are displayed. I do do like the interface for nVidia Inspector and will probably try that one first.

http://www.guru3d.com/files-details/nvidia-inspector-download.html

The problem is that the dungeon farms were simply too rewarding in relation to the rest of the content, though I think the nerf should have been more like 50%, not 66%. 33s per path is not great, though, 50s I think is a little more reasonable.

If you want to make money, SELL EVERYTHING you don’t think you’ll need and never walk by a harvesting node (You get sellable materials AND XP from harvesting). If you don’t plan to make your own food, why sit on hundreds of raw material items in your bank, same goes for gems and the like. I personally never salvage anymore, I don’t need any more mythril, and luck is a lie. If using the TP, sell for at or above the current available and when you need something put a low offer in. This is ESPECIALLY true for very high volume items such as raw materials. You may wait longer for stuff, but ultimately make more and spend less. The TP is not a huge mystery and doesn’t have to be hours spent flipping to make a good impact on the gold in your wallet.

abomally.2694:

The heatsink on your card may be clogged with dust – right at the back of the card, where the air blows out. You’ll need to pull the cowling off the card to be able to see it.

It’s a fairly common issue if you’ve had the card a few years (it happens on laptops too).

Should be pretty clear, I tend to clean the whole thing with a compressor every few months.

RedZebra.2345:

edit: ah missed that from agnes, i would start lowering it with 20% see if it stable, and set it a bit higher again so trial and error

Both GPU and shader, or just start with lowering the GPU?

Agnes.3259:

My card is Gigabyte GeForce GTX260 Core216 – factory clocks (GPU/Shader/Mem) are 650/1400/2000 Mhz. Lowered it to 630/1350/2000 and game never crashed again since.

How did you come up with what to lower it by?

RedZebra.2345:

Could not hurt to try that and i hope it works for you.
But still 85 c after 6 minutes is rather high for a single card, i should expect stable below 80 C with ocd factory settings, normally they dimensioned the vent and cooling element in this way.
ex. for reference https://www.youtube.com/watch?v=nc3AZ6-1ios
Not having issues with other games says nothing, did y already monitor y gpu during gw2 play ? ex, TechpowerUp GPU-Z (you can make a log file, you will see gpu load, mem usage, temp etc.) ex. a bad pasting, bad fan settings,.. can lead even with low load to high temps. 100% gpu usage shouldn’t also be a issue for a card it should be designed for that, it only becomes an issue when something is wrong. Again stress tests say nothing when only run for 6 minutes, you desire a stable system at his highest capabilities. With nvidiainspector you can also change clock settings ( lowering them wont hurt y card and it is also cheap ).

Ya, I would expect it to be hotter. From what I’ve read, the eVGA cooling for the 600 series wasn’t great. The area where the computer is is also quite warm, especially in the evenings, so the card starts off hot over 50C standing still (4095MB NVIDIA GeForce GTX 670 (EVGA): 54 °C). The guy in the video started about 10 degrees cooler than me. My machine may need to be blown out again too, but I just did it over the summer, so it shouldn’t be too bad. I’ve heard the same thing about the 900 series, that eVGA cooling isn’t up to par as compared to most of the other makers and tends to burn hotter. I was leaning toward the MSI card (970/980) when I was looking at maybe replacing the 670. That said, overall, standing still things are pretty cool, with the CPU at just 31C. I looked at GPU-Z last year and nothing jumped out at me or anyone I spoke to, but will do it again.

The ONLY way to cause the crash is to play GW2. I’ve found nothing else, and no other way to cause it. If I leave the shadows/shaders/post-processing on high settings the crash will usually come within 15-45 minutes, if not sooner. You say that other games don’t matter, and says nothing, but it says lots. It tells me that it’s likely some issue where the GW2 client is in some way implementing the video that the driver does not like, or, more likely, that it’s related to something the GW2 client is doing that my other games do not (Network, maybe…framerate spikes/dips causing issues on the bus maybe). Either way, the root is GW2, that’s what the other games tell me. Perhaps GW2 is causing the card to overheat (Points to a poor implementation of the video, since it’s certainly not graphically heavy), this has been checked in the past, but I’ll check again. I’ve heard of driver crashes caused by framerate spikes, but I have the frame limiter set to 60, my crashes do increase if I leave it unlimited, however.

It would be nice if it ultimately turned out to be heat as I have tons of room for more fans and that huge case I have, though it has never been a problem in the past.

I’ve downloaded nVidia Inspector and will try the clock speed change as well.

Seera.5916:

Not necessarily. Depends on how many times the account was shared and what was done during that sharing. The other person asking for account back after sharing it may have had it shared a lot more than tyu. Each situation is different and what is decided for one instance is not necessarily the same for another instance.

Account sharing is against the TOS. What more needs to be said. If you violate the TOS you are at the mercy of ANet. You can plead your case, but if you violate the TOS and whether it is done knowingly or unknowingly, ANet has the right to step in and intervene and ignorance is no excuse. GW2 is not so expensive that account sharing was ever warranted or needed and it quickly becomes abuse if you’re sharing with someone on the other side of the world. With so much account bound these days, especially masteries, any account sharing is essentially cheating.

If you’re doing such things, it’s not a matter of being clever, it’s not a matter of fairness and only a matter of time before getting caught. It’s not a democracy, it’s not even a monarchy, ANet is an all-seeing God who can look in to the private accounts of each and every player. Like any God, they can choose to be merciful or enact swift judgement. If you want to keep your account play within the rules (Look them up if you don’t know them) and don’t anger the gods.

RedZebra.2345:

Leamas, furmark is a stress test, normally your videocard should become stable at a certain temp after a while at 99- 100% gpu load ( Stress is NOT 6 minutes) You can follow this also with nvidea inspector or another program. Mine for instance has 82 C at 1920×1080 (20% oc), 73C default clocks. If you temp is not yet stable at 85c you can not tell if your gpu is capable running at full load at y current settings. Keep in mind furmark only stresses the gpu unit, you won’t know for instance what happens if both cpu and gpu are stressed ex. is the power supply capable enough to sustain the needed current. In my case i have to reduce oc in 64bit client, as it seems the power drawn by the cpu in this 64bit client is higher then the 32bit client for my pc, and as i use my power supply at his limit i have to reduce the power drain somewhere.

edit : did you already try lowering video card clocks ? : strange colors, not responding driver, 85 c after 6 min?

No, I have not lowered the video card clocks or adjusted voltage to the video card. It is factory overclocked from eVGA. Unfortunately the version of eVGA PrecisionX for my card doesn’t like Windows 10 and MSI Afterburner, I’m not sure what the percentage numbers mean so I’m bemoaned to change anything. Can’t find any decent guides for Afterburner.

Realbench does stress test GPU/CPU at the same time and I’ve run this test successfully several times with differing amounts of RAM selected when I still had the 48G installed. I have a 1200W power supply and it has been stability tested by the shop twice, so voltage availability should not be an issue unless it’s set low in the card.

The problem I have with Furmark and the eVGA stress test utility is that using the burn-in test you have a real risk of damaging your card, even if it starts off good and to be honest, I’ve never been given a reason to think the physical video card is bad as I’ve never had issues with any other games or graphics utility, most of which with heavier graphics requirements than GW2.

My point is though, that if it were a load issue, even short tests under extreme load should almost immediately cause crashing issues and I’ve run the Furmark and eVGA tests dozens of times over the last 3 years. GW2 barely places any load on the CPU/GPU, in comparison. Ran them much longer last year. It just seems overkill to keep running the same tests only to get the same results.

It all still leads me to believe something else is causing faults that makes the video driver unhappy. Makes me immediately think of the on-board network since it only happens in GW2 for me and many of the games I see people having issues with also have required persistent Internet connections. I’ve seen lots of complaints about Witcher 3, GTA5, Diablo 3, Starcraft 2 and pretty much any MMO. You see complaints about “normal” game far less often. So, I’ve ordered a PCI (As opposed to PCIe) network card from Amazon. This should hopefully get the network traffic off the PCIe bus and remove any interference it may or may not be causing and for $15 it’s a cheap test.

The whole thing reminds me of the pre-Windows XP days, when you had a limited amount of IRQs and IRQ conflicts would commonly cause such instabilities.

I apparently had memory with mismatched timing installed. Some memory was replaced under warranty last year. I really should give the shop kitten about this, their techs should know better. So I removed the 32G of Patriot Viper memory and left in the newer 16G of Kingston Hyper-X. Still crashes in GW2.

It’s been an issue with GW2 for a long time. I think the first time I reported it to ANet/nVidia/eVGA was Jan-Feb of 2013.

I can explain what’s happening with Windows in the background, with me anyway, but not what the root cause is. Windows has a built in feature called Timeout Detection and Recovery (TDR) that is meant to prevent hard crashing failures. It basically expects each of the drivers to report in at set intervals (The default is every 5 seconds). If the driver does not report in, after a given number of failures it assumes the driver is hung and issues a reset to kill and restart the driver, hence the message “Display driver driver name stopped responding and has successfully recovered.” I assume it’s the same for both nVidia and ATI. For me two dump files are typically created, one showing the DirectX kernel killing the nVidia driver, a second showing the driver recovery.

I’ve tried increasing the TDR timeout from 5 seconds to 20 seconds and have still had crashes. Last night I had a hard crash to an instantly frozen desktop. No dump files were generated, no system logs were generated. This is about as severe as a Windows crash gets without corrupting the system. I’ve since completely disabled TDR check for the video driver, but don’t expect it to help much. I expect more of these severe crashes and less of those where it recovers (Because I’ve told it not to recover), but I want to “see” it happen.

The problem is, that Windows is doing what it is supposed to be doing and driver is probably hung, but for those who have the problem what is unknown is what causes the driver to hang in the first place? Why does it affect only certain applications, but not others? For me, in the three years I’ve had this computer, I’ve only ever seen this problem in GW2 and never in any other game that I remember.

The most common causes are bad memory, bad power supply or bad video card but pretty much any piece of faulty hardware can cause it. Other possible causes include bad overclocking and insufficient cooling (overheating). Less common, but possible causes could include bad peripherals such as keyboard, mouse or external USB attached hard drives, which may also point to a bad USB bus and ultimately the motherboard. Certain combination of hardware sometimes just don’t like each other. Some software is known to cause issues, but I’m not going to get in to those.

The problem is that most of these things you’d expect to affect many or all applications globally on the machine, but it doesn’t. Again, for me, it only affects GW2. One possibility that I still need to test is the possibility that the network interface on the motherboard is bad. This could explain why it affects only GW2 for me since I don’t play any other games which require a persistent Internet connection like this. Any faults thrown by the network adapter may only surface when the machine is under load. One test I could try, which I only just thought of, would be to download something while running a CPU/GPU stress test (Furmark, Realbeanch, or something similar) at the same time and see if the error can be reproduced that way. Though, I suspect I will have no issues.

So, the bottom line is that no one knows what causes it and it’s incredibly frustrating for those it affects. The crashes are only a symptom of a much more hidden issue. It’s like the machine is cursed and the only way to fix it is to replace the machine.

It’s been an issue with GW2 for a long time. I think the first time I reported it to ANet/nVidia/eVGA was Jan-Feb of 2013.

I can explain what’s happening with Windows in the background, with me anyway, but not what the root cause is. Windows has a built in feature called Timeout Detection and Recovery (TDR) that is meant to prevent hard crashing failures. It basically expects each of the drivers to report in at set intervals (The default is every 5 seconds). If the driver does not report in, after a given number of failures it assumes the driver is hung and issues a reset to kill and restart the driver, hence the message “Display driver driver name stopped responding and has successfully recovered.” I assume it’s the same for both nVidia and ATI. For me two dump files are typically created, one showing the DirectX kernel killing the nVidia driver, a second showing the driver recovery.

I’ve tried increasing the TDR timeout from 5 seconds to 20 seconds and have still had crashes. Last night I had a hard crash to an instantly frozen desktop. No dump files were generated, no system logs were generated. This is about as severe as a Windows crash gets without corrupting the system. I’ve since completely disabled TDR check for the video driver, but don’t expect it to help much. I expect more of these severe crashes and less of those where it recovers (Because I’ve told it not to recover), but I want to “see” it happen.

The problem is, that Windows is doing what it is supposed to be doing and driver is probably hung, but for those who have the problem what is unknown is what causes the driver to hang in the first place? Why does it affect only certain applications, but not others? For me, in the three years I’ve had this computer, I’ve only ever seen this problem in GW2 and never in any other game that I remember.

The most common causes are bad memory, bad power supply or bad video card but pretty much any piece of faulty hardware can cause it. Other possible causes include bad overclocking and insufficient cooling (overheating). Less common, but possible causes could include bad peripherals such as keyboard, mouse or external USB attached hard drives, which may also point to a bad USB bus and ultimately the motherboard. Certain combination of hardware sometimes just don’t like each other. Some software is known to cause issues, but I’m not going to get in to those.

The problem is that most of these things you’d expect to affect many or all applications globally on the machine, but it doesn’t. Again, for me, it only affects GW2. One possibility that I still need to test is the possibility that the network interface on the motherboard is bad. This could explain why it affects only GW2 for me since I don’t play any other games which require a persistent Internet connection like this. Any faults thrown by the network adapter may only surface when the machine is under load. One test I could try, which I only just thought of, would be to download something while running a CPU/GPU stress test (Furmark, Realbeanch, or something similar) at the same time and see if the error can be reproduced that way. Though, I suspect I will have no issues.

So, the bottom line is that no one knows what causes it and it’s incredibly frustrating for those it affects. The crashes are only a symptom of a much more hidden issue. It’s like the machine is cursed and the only way to fix it is to replace the machine.

I completed Verdant Bring the other day (Yes, I’m slow) but I did not receive any rewards for completion. I’ve heard there was a bug blocking map rewards in the new area for characters who already have map completion in the core game? Is this the case? Is it something ANet is working on?

I understand that people’s accounts get hacked, but my question is if you are an active player how could you not know. If you try to log in twice, the client kicks you out and tells you if you’ve logged in from another location (This I’ve tested). Even if it didn’t wouldn’t you notice your assets, such as XP, gold or resources changing or am I missing something here? I’ve been playing 3 year and always know approximately how much of everything I have.

Qugi.2653:

Kraiz.6720:

Chris Cleary.8017:

tyu.9470:

I am sure you’ll agree that false positives happen EVERY time you do ban waves no matter how much you try to check.

Cheating on 2 of your 9 accounts got all 9 of them banned.

9 accs……. I fail to see any real reason any normal player would need more than 1 account.

There is some person I forget name that brag all time about 30 account. They do for log in reward. Account easy make 40-50 gold month from use laurel for T-6 mat to sell on TP. When there was sale on account some people buy many. I do not think worth money but some think it is.

I fail to see how is this a cheat though, if they’ve legitimately purchased 30 accounts…who cares. On sale, it’s cheaper than buying 6 character slots. They’re essentially their own guild. I’ve often thought of buying my kids their own accounts, then we can actually play together rather than them always using my account. Physically it’s no different and would appear the same on the ANet side…multiple logins coming from the same IP.

Where I see it becoming an exploit is with the free client for the base game, if you still get daily login chest, one could go out and sign up for a pile of free e-mail address and create a bunch of accounts that way, then siphon the funds to the main account via in-game mail. Agian, very hard to police.

Recent changes that still require in-game testing:

• Again, cleaned out the nVidia software using DDU and reinstalled the recommended 347.88 driver (Recommended version from nVidia for cards 970 and under), immediately disable all services to prevent further update. Windows 10 is also set to not automatically update hardware drivers, though it is telling me there’s one available.
• I lowered my base clock from 100.02 to 99.76. Figured what could it hurt. Not going to fry anything lowering the clock speed.
• After changing the clock speed, I reran the Intel CPU Diagnostic Utility and reran the Furkmark stress test (1920×1080) for several minutes (I shut it down when the card reached 85C, after about 6 minutes, as I don’t want to risk damaging the card). I also downloaded the ASUS Realbench utility and ran that stress test for 15 minutes twice. Once for 8G of RAM and a second time for 32G of RAM. All tests came up clean and received a “healthy” status. Mind you I had no problems with these tests before changing the clock speed either, but I more wanted to ensure I didn’t introduce any obvious instabilities.

Changed the topic to be more representative of what people see, rather than the technical name

Check your Windows Event log for system right after the crash see what Windows thinks is causing it. Do you have an nVidia card? Is the crash related to nvlddmkm? Typically, the event will say “Display driver nvlddmkm stopped responding and has successfully recovered.”. If you’re getting this error, I can suggest some settings that you can use to reduce the crashing, but there’s no real cure for it that I’ve found in 3 years of playing GW2.

If it is that error, check here for settings that might (Or may not) help…(I apologize for it being a bit long winded.)
https://forum-en.gw2archive.eu/forum/support/support/The-dreaded-nvlddmkm-event-log-error/first#post5724404

Check your Windows Event log for system right after the crash see what Windows thinks is causing it. Do you have an nVidia card? Is the crash related to nvlddmkm? Typically, the event will say “Display driver nvlddmkm stopped responding and has successfully recovered.”. If you’re getting this error, I can suggest some settings that you can use to reduce the crashing, but there’s no real cure for it that I’ve found in 3 years of playing GW2.

If it is that error, check here for settings…(I apologize for it being a bit long winded.)
https://forum-en.gw2archive.eu/forum/support/support/The-dreaded-nvlddmkm-event-log-error/first#post5724404

NitricRose.1863:

Was working fine for a long time and then my computer crashed hard.
Windows 10 will no longer boot. No ideas why.
Boots linux fine (gw2 linux port?).
Now I’ve got a long messy process of rebuilding my system. Bleah.

I’m going to pass on testing this any further. This kind of bad luck I don’t need.

Check your boot order for the drives. Sometimes a hard crash will screw up the boot order in the BIOS and it doesn’t find the Windows drive. I’ve had this happen a couple times.

lynspottery.6529:

Johan Skold.4708:

Leamas.5803:

Has the 32-bit client and OOM been linked to the nVidia driver crash or will the new client not help with that?

*Display driver nvlddmkm stopped responding and has successfully recovered.

The description for Event ID 13 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer. *

That sounds unrelated. If it happens again you should try downloading new nVidia drivers.

I just checked on my nVidia selection from the program management box and you can add the new .exe file to the list of items it will work with.

But on another note Johan, it appears that mumble overlay for chat is not recognized with the new client. Is there a fix or work around for this? I would hate to have to revert to the 32bit just to be able to voice chat with my guild members.

Ya, you’re right. My current nVidia settings are for the x86 client. The only difference though, should be the Shader Cache, which will be disabled for the x86 client, but not for the x64 client. I didn’t see a need to disable the shader cache for other games. Even with GW2, it’s more of an experiment. The power management change and multi-display mode change I made at the global level.

See the following post for my solution and summary. Read through the thread for history and what was tried.

https://forum-en.gw2archive.eu/forum/support/support/The-dreaded-black-screen-of-death/5755714

Well here’s my 3rd annual attempt/rant to try and resolve this issue. It’s a little disheartening that it has never been resolved and that nVidia still really hasn’t been able to include something in the error messages that points to the actual issue.

The error looks like this in the Windows event log:

_Display driver nvlddmkm stopped responding and has successfully recovered.

The description for Event ID 13 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

If the event originated on another computer, the display information had to be saved with the event._

I’ve done many of the things suggested in this, and many other forums, to no avail. Some things “improve” the situation, but nothing fixes. Like with many others this problem only affects a single game, in my case Guild Wars 2, though I think I remember it happening in Bioshock Infinite, but it was no where near as persistent or invasive as with GW2, where it will often completely freeze the machine and sometimes send it for a complete reboot. Could be completely unrelated crash scenario too.

This was all done this time last year and much of it the year before that.

• memtest86 – Found 3 of 8 memory sticks defective. The store replaced the one bad one I bought from them and I still have to send the others back. So now I’m running 48g of RAM rather than 64g.
• Furmark burn in test – Card gets hot but seems to settle in around 81C with a fan speed of ~2600RPM. Well within safe range
• eVGA stress test utility – Similar to Furmark, similar results
• Intel CPU Diagnostic Utility – Came up clean
• PSU – I’ve had the shop I bought it at test the PSU twice and both times it came up clean.
• Drivers – It happens on all driver versions I’ve tested from 297.xx right up to the most recent. 306.97 was the most stable for GW2 (Though it still crashed periodically) until it’s age started causing all kinds of compatibility issues with other things. For a while, what I was doing was installed 306.97 to play GW2 and upgrading it to play other things, but that’s really not reasonable.
• Windows version – It was happening on Windows 7 and now most recently Windows 10, using the most recent driver as of a week ago.
• nvlddmkm.sys – Some claim that driver updates do not update this file properly. So I used DDU, and confirmed the file was gone before installing the new driver, only selecting the driver itself and PhysX. I have noticed nVidia User Experience is still in the task manager even though I told it not to install it. I need to confirm Windows automatic driver setting and redo this step.
• Kill Apps – I have killed all non-essential startup apps with the exception of Steam. When I play GW2, I close Steam, and through task manager kill any Adobe Services, any nVidia services (Including user experience), any Apple Services, and anything else that doesn’t look necessary.
• Hardware Peripherals – I’ve purchased a new monitor (Well, last year, Asus VG278he) and have swapped to different keyboards, which I have several spares of the same (Fujitsu FKB4725 mechanical PS2 keyboard) and mice (Standard MS 2 button laser mouse). I’ve unplugged all external hard drives and the printer.
• HWMonitor – It reports invalid voltages, which is not uncommon for this tool, the BIOS says voltages are fine. HWMonitor says core temps are all fine, below 50C.
• MB BIOS is at the most current version
• Fast boot is disabled and Windows 8 power option is set to “Other OS”
• Downgraded driver to 347.88 and did a clean install using DDU to remove the previous driver.

What helps? What doesn’t?

• In Windows, setting the power profile to high performance for both the machine and the video card, rather than adaptive does help
• In GW2, setting the frame limiter to 60fps rather than unlimited helps. Driver 310.14 (Or was it 314.10…) was unusable if the frame limiter was not set to 60. The game would crash within seconds. Went back to 306.97 at the time.
• None of the other graphics option in GW2 seems to make a difference in how often the crashes happen.
• Where the other graphics option come in to play are the higher the settings, the more likely the crash is to require a hard reset or to cause the machine to spontaneously reboot. Different graphics setting can also change the color of the crash screen from black to red to white or this morning when I was testing shadows on ultra, shaders on high and post-processing on high I had orange, then the machine rebooted itself.
• Keeping the setting low, typically, the driver will crash to a black screen, the keyboard and mouse will become unresponsive, the background audio will continue as normal for the duration, then the screen, keyboard and mouse will come back after 10-15 seconds (At which point, I’ve typically died in game). About 1 out of every 4 crashes will require a hard reset in this scenario.
• I’ve not managed to nail down anything specific in game that will trigger the crash. It can happen anywhere, doing anything, even just standing around chatting
• Occasionally, twice it’s happened, the crash is so severe that it screws up the boot order in the BIOS and I have to go in and fix it.

Computer is:
Intel 3930k CPU (Standard factory clock speeds)
Gigabyte X79-UP4 (v1.0) Motherboard
eVGA nVidia geForce 670gtx OC Edition
48G RAM Patriot Viper (Was originally 64g)
Boot drive 240g Sandisk SSD (GW2 is installed on the SSD)
Other drives – 4 internal conventional drives. (One could potentially be removed, the others are needed)
LG BD-ROM
Corsair Professional Series Gold AX1200 Modular Power Supply
Case ThermalTake Level 10-GT (Cable routing is all very neat and tidy)

Other things to try:

• Check paging file size and increase to at least 16G if it’s smaller. (DONE)
• I found a forum entry that suggested adding this to the registry, though I’m getting deja vu here, so I think I might have done it last year to no avail. [HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Contro l\GraphicsDrivers]
  “TdrDelay”=dword:00000008
• I found this, that I have to check in to yet, but could it cause issues? If your base clock is above 100 put this to 99.76. Some claim it a known bug in gigabyte bios’ (DONE)
• Still need to download and run Speccy, which I will run after killing everything as I still get crashes in that state.
• Still need to check on automatic updates and ensure they’re disabled:
  http://winsupersite.com/windows-10/stop-automatic-driver-updates-windows-10
  https://support.microsoft.com/en-us/kb/3073930 (DONE)
• Look again for a tool that will monitor memory, cache, page file usages, actual voltages (NOT The bogus HWMonitor ones) and CPU/GPU core temps, etc in real time and write to a file every few seconds. This should point to the appropriate area if the driver is freezing while trying to write to a certain memory/cache location because it full. Previous attempts to find such a tool were fruitless.
• Maybe increase voltage to the GPU though this seems unlikely to work. If it were a load/voltage problem the Furmark burn in test would surely trigger it. GW2 doesn’t even make the card break a sweat. Skyrim with all HD textures has never crashed.
• ArenaNet has a 64-bit client in beta. I’ll have to try this and see if the 64-bit client vs the 32-bit client helps the issues. (DONE – But too early to tell if it will make any difference)

ArenaNet says it’s not their problem since not everyone using nVidia has issues. eVGA/nVidia tend to blame GW2 since that’s the primary source of the issues, though, again NOT EVERYONE USING NVIDIA HAS THE PROBLEM, which indicates GW2 is not the sole issue, and I’ve seen LOTS of people having the identical issue with all levels of cards ranging from 400 series, right up to the Titans, in other games, such as Witcher 3, GTA5 and even older games like Starcraft 2. Searching Google you see this issue going back probably 10+ years. The place I bought the computer, and I’ve brought it to them twice while the machine was still under warranty, says it must be software/driver related since they can’t reproduce the issue and get no errors using their stress tests. No one will take responsibility and no one has ever been able to help. It’s all very maddening considering it was a $3000+ computer at the time and has never worked right from day one.

I don’t want to buy a new card (At ~$450), as I have no faith or guarantee that it will help given that the new cards seem to be just as plagued with the problem as the old ones and I see many people RMAing or otherwise replacing their cards only to have the same problems with the replacement, which points to a computer hardware or a Windows software problem or compatibility issue. But, again, where to look when all stress tests and hardware checks pass with flying colors and more importantly, HOW to look? What tools will do this kind of real time monitoring/file logging to the hard or network drive in the background. My guess is there’s a faulty cache somewhere in the hardware that the nVidia driver uses, that regular stress tests don’t identify. We all know nvlddmkm.sys causes the crashes, but what DID it do to trigger it or what could it NOT do that caused it to become unresponsive or otherwise unstable? It points to GW2 using some nVidia driver function that not everyone uses that is using some resource that is broken or faulty or not supported on the given machine, but again, how do we test this or pin-point it?

This wall of text was brought to you by me and is the combination of 3 years of trying to resolve this issue. Yes, I’m THAT stubborn.

Maybe something here will help someone, maybe someone has some other suggestions I’ve not yet tried. Either way…here’s is my annual post regarding this issue.

Are you running SLI? Doesn’t appear so, GW2 runs like a dog with SLI, at least it used to? GW2 tends to have “unique” issues for some of us lucky few. Does it do the same thing with the 32-bit client?

Nick.8256:

For years I have been going back and forth with support trying to fix these oom crashes and this has fixed everything even with massive overclocks (like i suggested very long ago). Anyways, if you guys are still crashing I would suggest giving asus realbench a shot to test for system stability. Its pretty cool because unlike other tests it will test everything at the same time. Set it to stress test mode, select your installed ram size and let it run overnight for at least eight hours.

http://rog.asus.com/376212014/overclocking/realbench-v2-4-launched-with-x99-support/

Can this be used for non-Asus motherboards?

Lunacy Solacio.6514:

Only other things I could see, is to check the event viewer, and the memory dump from the computer crash. I would also suggest running memtest from usb/cd boot. It’s entirely possible something is faulty. Assuming no overclocks? It’s possible the power supply isn’t up to par as well… in fact many manufacturers skimp on this.

also, make sure everything is backed up that you want to keep (good precaution anyways, but so few of us actually stick to this and it hurts when you lose things you want to keep >.> )

After running memtest, if it runs for at least overnight without issue, then that may be ok, I’d move on to making sure chipset drivers are up to date from the actual motherboard manufacturer. If were able to update them, try the current graphics drivers again. If this isn’t working, reinstalling windows is the next step. If you have tried that before as well, and error persisted, there is something wrong with your hardware somewhere x.x

It was a custom build, not a pre-built, so nothing was skimped on. 1200W power supply, 64G of RAM originally (48 installed now), nVidia 670 which was a $600 card at the time. Every component 4-star rated or better.

I’ve done most of this before. memtest86 was done a year ago and found 3 of 8 bad memory sticks were bad. Two was replaced at the time (Leaving me with 6 total). Didn’t help with the crashing though. I’ll have to check again. Still haven’t gotten around to RMAing the other 2. PSU (1200W Corsair Pro modular) was tested twice by the shop I bought it at and came up clean. Furmark burn in stress test will run indefinitely with no crashes and settles in with a temp at about 81C. eVGA stress tester also runs fine so I think the physical GPU is fine, or at least the parts the stress testers test. Windows has been reinstalled to no avail. I updated chipset about a year ago on Windows 7, the last time I tried to resolve things, but will try again with new Windows 10 drivers I downloaded this morning. This time last year, I took it in for one last attempt to get it fixed under warranty and the shop did a week-long stress test, which I paid $80 for, and found no hardware issues. It’s all really kind of maddening. LOL

The New x64 client
The new client changed the type of crash from a system crash to more of a CTD (I forgot to check the event log though to see if it was caused by the nVidia driver). The two times it did crash, my desktop came up and I had to fiddle around with task manager until GW2 was in a state that it could be killed. So far it seems to be an improvement though I’ve changed a lot of settings too and reverted my driver. But up until now I’ve never had the software crash, always a system crash, no matter what settings or driver I used, so we’ll see. The only difference the setting and drivers made was the frequency and severity of the crash. Right now I’m running pretty nerfed with shadows turned off and the shadow cache in the nVidia driver is turned off for GW2 (Figured I’d try this), shaders are set to low, post processing is set to low, Light adaptation is turned off and the frame limiter is locked at 60fps. My memory in the BIOS is set to stability mode (At the sacrifice to performance).

Znake.9762:

Leamas.5803:

Thanks you!!!
I’ve only been asking for this for 2 years. Hopefully, this will clear up the infamous event log message. This is the last thing that gets written before GW2 crashes.

*Display driver nvlddmkm stopped responding and has successfully recovered.

The description for Event ID 13 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer. *

These days, I’m so pessimistic about it after fighting with, Anet, nVidia, eVGA and the people I bought the computer from for three years, that deep down I’ve convinced myself there’s no way it’s going to work and that the only way to absolutely get rid of the error and hopefully the crashing it is to replace the computer and use ATi.

As that message says, it’s a problem with the display driver and getting a new game client won’t help with that at all. In my personal experience, even a reinstallation of Windows can solve such problems but it’s very hard if not impossible to find out the real reason for those errors sometimes.

The New x64 client
The new client changed the type of crash to more of a CTD (I forgot to check the event log though to see if it was caused by the nVidia driver). The two times it did crash, my desktop came up and I had to fiddle around with task manager until GW2 was in a state that it could be killed. So far it seems to be an improvement though I’ve changed a lot of settings too and reverted my driver. But up until now I’ve never had the software crash, always a system crash, no matter what settings or driver I used, so we’ll see. The only difference the setting and drivers made was the frequency and severity of the crash. Right now I’m running pretty nerfed with shadows turned off and the shadow cache in the nVidia driver is turned off for GW2 (Figured I’d try this), shaders are set to low, post processing is set to low, Light adaptation is turned off and the frame limiter is locked at 60fps. My memory in the BIOS is set to stability mode (At the sacrifice to performance).

nvlddmkm error
Finding the real reason is a real pain. There doesn’t seem to be a tool that will monitor the resources that need to be monitored in such a way that it will write out to a file right up to the point of the crash. I’m convinced it’s a bad cache somewhere in the hardware that’s getting used in some special way by GW2 (via the nVidia driver) that most other games do not. Unless I can find something that will monitor such things and write the results to the HD (Or network drive), ya, finding the solution may ultimately be a piece by piece HW replacement or a full PC replacement, though I see LOTS of people with brand new PCs who have this problem (Including myself at the time I bought it). One thing I did notice is that many of the games that experience this problem tend to have heavy on-line requirements. I wonder if the NIC might be throwing faults causing the video driver to freeze up since it shares the same bus as the video card. Again very hard to confirm.

deltaconnected.4859:

Whatever you do just avoid the r355 branch and newer. Doubly so if you’re on SLI (it’s completely broken for it). Ideally 347.88 as that was the last time I had no NVidia-related crashes, and 353.49 if you’re on 980ti’s because it’s the newest driver that doesn’t suck for performance.

Why though? What is the instability? What typically causes it? Why does it work fine for some and not for others? These are question that no one EVER seems to be able to answer. ANet blames the hardware, nVidia/eVGA blames the software or other non-GPU hardware, the techs where I bought the computer blames ANet because all their test show the hardware is clean…it’s a vicious circle while my computer continues to crash along.

The biggest question is Will the x64 version of the client help? You say, “no”, but what does ANet say? Can’t make it much worse I guess since I typically crash every 15 minutes or so.

On my old machine (First gen i7) I had a pair of 275s in SLI. In general everything was fine for most games, but GW2 HATED it. I had to turn the SLI off then GW2 was happy to run on a single 275 card, lower graphics settings of course, but at least it never crashed. Got the new machine with the new 670 (3 years ago) and GW2 started being a dog.

deltaconnected.4859:

The “something” is just “attempted NVidia optimizations” lol. Nothing to do with GW itself. I’d list how painful surround has been as an experience (start menu far left on win10 since TPs, can’t disable surround to try non-surround without reinstalling, TW3 more broken with no fix in sight that makes the GW experience a pleasure, which says quite a bit given SLI+surround is broken for the newer drivers for it, among other less noticeable things) but it’d take up more than this page sooo…

Nahh, attempted NVidia optimizations is not it…the driver crash issue has been a problem with GW2 since at least driver version 306.97 and January of 2013, then it would get worse with newer drivers so I stuch with 306.97 for a LONG time, periodically trying newer ones, then it seemed to stabilize a bit with driver 332.xx I believe, but you can’t stay on ancient driver versions for the sake of a single application. GW2 is literally the ONLY game I have that I’ve ever seen this type of crash in. Ultimately I was forced to upgrade my driver again because ESO would not work on on the older driver and flip-flopping drivers is a PITA.

deltaconnected.4859:

If you’re getting nvlddmkm.sys crashes, try downgrading to 347.88 (980 and older) or 353.49 (980ti). That’s caused by something else so I doubt switching to x64 would help.

Ya, nothing helps much and no one has ever been able to pinpoint what that “something else” is or why the driver works fine for graphically heavy games like Watchdogs or Dishonored, but GW2 runs like a dog. It has been a persistent problem with nVidia hardware for over 10 years and it’s a friggin’ joke. It’s like a curse, once you have it, you never get rid of it.

This is something that the regular testing tools and benchmark tools do not touch. No problems indicated with Furmark or the eVGA stress tester, Intel CPU health check is clean, had the PSU checked twice at a computer shop, memtest86 is clean. It points to GW2 using some nVidia driver function that few other games use that is using some resource on the affected person’s machine that is broken, faulty or not there, again, that is not being used by most other games. 200+ games in my Steam library and this is the only one this happens to.

Still, I’m hoping the x64 client will help at least some.

You know, the irony is I JUST today posted a huge rant on nVidia’s forum about the crashing issue in GW2 and how it was linked to the nVidia driver.

ZudetGambeous.9573:

Well I downloaded the client and put it in the GW2 directory and ran it…

I have no idea what to do now. There isn’t any error code or anything, windows just says no

Your download looks incomplete, try redownloading the file. The 64-bit version is about 4m larger than the 32-bit one. Should be about 28M as opposed to 24M.

How would you like testing to be done? Max out the settings as they should be, if one has a card that will do it, and see what happens?

Has the 32-bit client and OOM been linked to the nVidia driver crash or will the new client not help with that?

*Display driver nvlddmkm stopped responding and has successfully recovered.

The description for Event ID 13 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer. *

Thanks you!!!
I’ve only been asking for this for 2 years. Hopefully, this will clear up the infamous event log message. This is the last thing that gets written before GW2 crashes.

*Display driver nvlddmkm stopped responding and has successfully recovered.

The description for Event ID 13 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer. *

These days, I’m so pessimistic about it after fighting with, Anet, nVidia, eVGA and the people I bought the computer from for three years, that deep down I’ve convinced myself there’s no way it’s going to work and that the only way to absolutely get rid of the error and hopefully the crashing it is to replace the computer and use ATi.

Treacherous Path gives you credit for Pact Airship Wreckage and vice versa. Finished my last one today. I’m sure it will be fixed in the next patch, but for now, at least they’re attainable.

From what I understand, they’re reversed, Treacherous Path gives you credit for Pact Airship Wreckage and vice versa, but when I did Pact Airship Wreckage, it didn’t tell me I completed anything and is not showing completed, but does tell me I learned all I could when I try it again immediately. I haven’t gotten back to Treacherous Path to try that one again, but I am still sitting at 10/11 done with Pact Airship Wreckage showing on the map as the only one not completed.

Hmmm I know, long ago, craft leveling was much better than it is now. Either way, it appears I was mistaken. I can admit when I’m wrong.

Ayrilana.1396:

How about long and involving quests for each of the tiers that take as long to do as earning the XP?

Masteries are part of the new progression and I see no reason to retroactively award progression to them based on previous XP earned. Would you expect other games to retroactively reward you with XP when they raise the level cap?

Level cap increases are evil. No, I don’t think previous XP earned should be rewarded retroactively, but existing content/functionality should not be taken away and/or arbitrarily locked behind gates. They seriously couldn’t come up with 5 things for the Pact Commander mastery track that didn’t affect core content? Seems short sighted to me.

Inculpatus cedo.9234:

Well, you could have obtained it free of charge, had you merely logged into the game (even for just 5 seconds) during each period the Episodes were active.

Sorry to hear you weren’t interested at the time things were given out free.

Good luck.

This. Even in the last few months, and over the summer, I simply didn’t have time to play, and still haven’t done the last episode of Season 2, but I did try to login as often as possible, if for no other reason, to unlock the content and get the daily login reward. Over that time, I’ve amassed 6 exotic armor chests (Sitting in the bank waiting for a new alt) and nearly 150 extra laurels after getting all my T6 to 250, essentially for the 10 seconds it takes to log in. It’s a helluva bonus and from ANet’s perspective, it keeps GW2 in the back of the mind of players on “vacation”, so ultimately more likely to return.

I’m OK with the mastery system in the new content, but I’m already bitter about core content masteries that I now need to grind out XP for, when I’ve already put in countless hours and rolled 80 hundreds of time, for what…for them to be converted to essentially useless spirit shards, of which I have nearly 900 (Other have far more to be sure). We should be able to spend the spirit shards to unlock core content/functionality now locked behind masteries, provided we have the mastery points to spend.

I did tailoring last night because I wanted Sinister gear for my necro. Sinister gear can only be crafted. So I figured I’d use the opportunity to level up a mesmer (Crafting is normally good for about 70 levels if I remember right), but to my disappointment, the new character sadly only gained 7 levels, plus another 2 crafting the Sinister gear after reaching 400. So, when all was said and done, she only went from 20 (Used a scroll to start at 20) to 29. I assume one requires “Productive Downtime” in order to get the proper amount of experience from crafting.

What a kick in the face for someone who has been playing the game since essentially launch.

Now it becomes a question of whether I even want to stick with this character or abandon her, and turn her in to another bag mule. I have enough to do with trying to work on the HoT masteries and elite specs with my existing 80s and working on a legendary at the same time. I don’t need the BS of leveling a new 80 from essentially scratch. It also used to cost 254000xp to roll an 80, now I’m looking at my account, while standing in Rata Sum, mastery rank 4, all in HoT, and it’s going to cost 635000xp to roll the XP bar.

I can hear all the criticisms now, about how I “expect” 80’s to be just given to me and to be honest I don’t care what anyone thinks., how it’s “only 2 level”, etc. After 3 years, and doing full map completion essentially twice (I know others have done it way more), I don’t play GW2 to grind through leveling 80s the hard way I play to play the content I enjoy. I realize that pushing out two masteries isn’t a huge amount of time, but core content, previously available, should not have been locked out for paid players. Only new things should have been added. It’s a load of crap!!

In my opinion, Pack Mentor should be the last one of the the Pact Commander masteries gained and the crafting one should not be there at all. Like the original commander tag, Mentor is too easy to get and will be used as a status symbol for those who have no business having it.

SoulstitchMMO.1396:

If PvE players are so hot for “progression” why not make masteries per character. Wouldn’t that be more rewarding for them when they are going for elite specs?

No…just no!!!

I don’t like Logan either. Kill Logan and bring back Spioler if you haven't done your personal story...Magister Sieran in some way. I'm sure the Pale Tree should be able to do that in some way. In my opinion, she was one of the best acted and portrayed NPCs in the game.

Won’t happen, most of the maps rely on gravity to keep you where you have to be. It would be an enormous undertaking, with no gain, to lock down the maps and complete structures so the gliding doesn’t break the core game and doesn’t look stupid. Hell, they haven’t even completely locked down the maps for regular play yet. I know of 2 or 3 places where you still can get either outside the map or in to incomplete areas.

ScribeTheMad.7614:

symke.3105:

I think they put too much emphasis, PR and marketing on new skill lines. Core Gw2 never had that feeling of “I have to buy this game, so I can play this or that class”. It was (and probably still is) about the fabulous world and plethora of things to do in it. With whatever class and/or skills you chose.

I’d say for a lot of us this is how we continue to feel.

The class is the vehicle by which we experience the world.

Experiencing the world is what we want to do.
Some of us like to experience driving different vehicles.

But when you tell us our range rover is getting a sunroof, an engine upgrade, and new offroad tires when they open up a new territory to explore our thought is “sweet gunna go experience more of the world with my upgraded ride”

We don’t expect to have to drive 10,000 miles in the new territory with the upgrades installed before we’re allowed to open the sunroof or go offroad.
(we don’t mind getting some experience/learning how to drive offroad, just not that much)

See, I didn’t expect anything more than what we got. ANet does not have a great history of making things easy to get or free. In fact I was considering the elite spec basically as the reward for actually playing the HoT content, much like unlocking all the core specs is essentially the reward for completing leveling to 80 in the base game. Everyone should be happy they didn’t convert the base game to the same model, where you actually have to play the game and do hero challenges to unlock your specs.

It’s not in their best interest to make things easily attainable…people would finish it quickly, get bored and leave.

ScribeTheMad.7614:

BulletNG.7069:

What is the conversion ratio from Skill Points to Hero Points? My Warrior had well over 400, if not 500 Skill Points before HoT, but I still don’t have enough Hero Points to unlock everything.

You earn a set number by lvl 80, all excess beyond that were converted to spirit shards a while back. (so that they could control how quickly we have access to stuff)

This makes sense. I knew I had more than what I started with.

I see lots still complaining even with 250…this is meant to be “elite” people. I don’t necessarily even think 400 was unreasonable, though I am happier with 250. That said, for all those with several alts, including myself, soon to have 4×80s and I know others have many more, you CANNOT expect to unlock elite specs on a necromancer and automagically expect them to be given to you for your warrior or mesmer. The argument about being blocked behind masteries falls flat since masteries are account bound and only have to be done once and all your alts can use them. Something that might be a workable compromise, and I mentioned this in another thread, is to have the elite spec profession bound that you only have to unlock it once per profession, but even that seems like a bit of a shortcut to me….why should it be free?