Let’s put it this way I have almost a 5 gigahertz overclock. And it doesn’t get to 90c playing wow
@OP - have you tried going into UEFI and disabling Intel SpeedStep and Intel Speed Shift? I was having the same problem and that seemed to work around the issue for now, but it does cost a good bit of FPS. . anywhere from 30-50% lower. No more crashes or Error 132 now though. . took the tip from the Blue post in the thread posted above:
Hope that will help! Also hope Blizzard fixes this ASAP!
I wouldn’t bother doing that, all you’re doing is throttling the power consumption. That’s why the Blue post suggested trying disabling turbo boost and multicore enhancements. A lot of motherboards have built in boosters. Like for instance: ASUS has APE(asus performance enhancer), which messes with things like the PL1 and PL2 durations. Most motherboard companies default these enhancers ON, since they know that most users aren’t tech savy and just want the fastest thing they can get their hands on.
An example: This PC I’m using has a 13600kf in it. It’s a 125w CPU, but can turbo boost up to ~200w. The duration it’s allowed to do it is normally baked into the CPU itself, but motherboards can override it. Since I have good liquid cooling on mine, I can run mine in perma-turboboost(PL2) mode and my CPU will never get above 75C, no matter how long I leave a torture test running and won’t drop out of turbo boost mode either. If I didn’t have that kind of cooling and was still using those performance enhancers, my CPU would spend 99.9% of it’s time in either the non-boost(PL1) mode or thermal-throttled mode during that kind of testing. This could lead into some weird issues if there’s some kind of imperfection in the BIOS or how the motherboard handles things like voltage regulation. It won’t harm the CPU at all, but there could be some irregularites that arise elsewhere in the system.
A few days ago, I also brought up these potential issues with the 13th gen and motherboard(BIOS) issues:
People having these issues with 13th gen CPUs, can you please specify which version you have, what motherboard you’re using and what BIOS version you’re on? An easy way to get the information is to press start → type system information → open it → on the first main page that opens up, you’ll see all the info you need.
I get your point, I monitor temps and they are in well within the normal range (nothing over 60C for either CPU or GPU no where near where thermal throttling would come into play), also stepping and timing come into play with over clocking as well, not just voltage. Besides I do not do any manual over clocking, or voltage mods. . everything is in it’s stock config. . there very well may be BIOS issues at play. . but the thing that stands out that makes it really hard to blame the hardware, is that I play about 15 or so other games that all are way harder on the hardware than WoW. . and WoW retail… . is the ONLY one this happens on and it seems others are having the same experience. And it seems to be only since 10.2 dropped for most, and for me it really only started this week.
Divide by zero could very well mean that instructions are executing in an order the code is not prepared to handle when SpeedStep and Speed Shift are enabled (i.e. stepping and timing).
It does seem from these 2 threads that a lot of them are on Intel i9 13900 variants.
Right now though as it stands. . . if you are plauged by this issue. . disabling those two options in UEFI may be the only way people can play until an RCA is executed and an actual cause is found.
The temps aren’t really the big issue, it’s the PL state(turbo boost) more than anything. Unless you changed the prochot or overheat shutdown protections, the CPU can run at it’s throttle temp indefinitely. The issue is when you have these motherboard performance boosters going while that’s happening(if your PC can’t hold a stable temp under constant turboboost load). It can lead into a square-wave like pattern similar to hitting the rev limiter on a car in neutral. THIS is likely what’s causing the problem for some people.
Voltage would be the main culprit. Much like what I was just talking about, constant transient spike loads from a CPU doing some power level yo-yoing could cause voltage sag in other voltages around the board; if the BIOS is wonky or there are hardware design issues on the board or if the power supply couldn’t handle it. This could then affect the ram and ram is ultra sensitive to voltage fluctations.
Yeah, that’s possible, but again, the ultimate end result of those features is basically doing the same as jumping between PL1 and PL2. Except they are just straight up messing with things like the clock mod or multipliers. Like taking it from a 51T down to a 38T or something. But CPUs are extremely complex beasts that account for all the timing variances like that, so I highly doubt this is some out of order execution issue.
I’ve seen it around the gaming world a lot lately, 13900 in particular. But again, that could just be like what I said earlier about it being a top of the line component and the types of people that would buy that are usually enthusiasts that are likely going to be trying to milk all the extra performance out of it that they can; using features like BIOS performance enhancers.
1 Like
I don’t disagree, but I’m not sold that’s 100% the case here. . like I said I run everything stock, so there is no speedstep or speed shift on top of an overclock and temps stay stable (my setup is very conservative), some other cases for hardcore modders maybe, but the other thread seemed to be all Alienware Aurora R15 owners (including me), which are prebuilt and not really the ‘modder’ audience.
Also, I don’t think it’s a coincidence that Blue has stated in the other thread that they’ve seen a recent uptick in Error 132 for 13th Gen CPUs and it seems to be exclusive to OverWatch2 and WoW Retail. The uptick is very very recent coinciding around recent updates. Look at the start dates on these threads one is 11/27 and the other is 11/29. 10.2 for WoW was released on 11/7 (with some hotfixes later than that) and OverWatch 2 had a major patch on 11/14.
The circumstantial evidence leans heavily to something changed in their code that is not playing well with i9 13th Gen CPUs when SpeedStep and Speed Shift are on.
I have an i9-13900KF and I think my BIOS version is Alienware 1.6.0 (I have an Alienware R15). What’s happening with me is I can’t run WoW unless I overclock my system, but I don’t think I can do that at the system level – I have to use the Alienware Command Center to overclock. But the Alienware Command Center is prone to crashing, so that tells the computer to stop overclocking… which then crashes WoW.
I found a fix (game hasn’t crashed in a couple hours anyway).
Go into the Intel Extreme Tuning Utility software and change Performance Core Ratio from 55x (or whatever your default is) to 54x or lower.
Edit: it didn’t COMPLETELY fix it, but the first crash happened about 6 hours in. I lowered it to 52x and I’ll see if that completely mitigates it
1 Like
For Alienware users be aware that Intel XTU and Alienware Command Center DO NOT like each other and Intel XTU will likely require you to uninstall and re-install AWCC to fix it if you wish to go back to it.
1 Like
Yeeeah… found this out the hard way earlier lol
Although, I do wonder if uninstalling AWCC and just not using it would fix my problem. But that’s something to think about for when I have more patience and optimism.
Oh, didn’t even see this. Definitely seems to be some kind of Dell issue. Maybe I’ll contact them for help.
Unlikely. . . worth a shot though. . I myself have it installed but only use it to manage the lighting. . I don’t use any of the overclock profiles and leave overclocking off, so that it runs at stock speeds.
Also from the other thread it is apparent the issue is not just for Alienware machines (others have the issue on non-Alienware machines). . . it looks like the common thread is actually the the Intel i9 13900 series processors when Intel SpeedStep and Speed Shift are on, and the Blue post there supports that theory.
1 Like
Hmm…
well, I have no desire to overclock if I can avoid it, so I’ll try disabling SpeedStep and Speed Shift in the BIOS.
i’ll be very happy if this works lol
Update: It… seems to be working? Usually my game’ll crash within a few mins, but it’s been stable so far. I’m very tentatively happy. Like, thank you so much – seriously. <3
Well this is the part I highly doubt. They wouldn’t code in some specific branch for a single model of a whole generation of CPU. The 13900 isn’t doing anything special over the 13700 or the 13600, as far as I know.
Which makes it sound more and more like a motherboard/BIOS issue.
You could also use a tool like ThrottleStop and just tick the disable turbo button, because that’s essentially what you’re doing by changing the ratio. Or from the Windows power profile settings, you could set the “maximum processor state” option to 99% and it won’t engage turbo anymore.
Doesn’t have to be for a specific branch, over the years I’ve seen different models of the same CPU family freak out over how things are coded on one model and be entirely sane on another model in that same family. You may be right, but if what you are saying is true. . . the ‘uptick of Error 132’ that Blizzard’s own team are reporting would not be recent. The reports would be statistically static (meaning no rise or fall in the number of reports). This post below however clearly states that there has been an increase in frequency in reports of Error 132. . . likely from their own internal reporting tools and/or from bug reports/posts. . .etc. .
An uptick in reports is a very strong indicator that something changed recently that has had an adverse reaction. Now. . I am making the assumption that the statement of multiple games is in reference to the reports of WoW Retail and OW2 crashing. In that case. . the likely culprit . . a recent software update ‘somewhere’ because that is the variable in this equation with the largest likelyhood of change. I would guess that they had a minor library framework update somewhere that shipped with the November updates. . and it slightly changed how some of the instruction sets are handled. . and they’ve stumbled upon a use case in live that their internal test cases and QA didn’t cover and capture.
As far as BIOS versions go. . . using the Alienware Aurora R15 crowd as an example. . version 1.6.0 of the BIOS that most of us are running? Came out on Sept 12th. So even that is a fairly static variable, given the timing of the Error 132 reports here. Hardware is actually likely the stable variable in this equation.
The reports in these 2 threads seem to say stability hasn’t been an issue until 10.2 for WoW Retail . . I’m not sure about OW2 I’d have to go look.
So if the hardware is fairly static and was stable prior to November software changes. . and not stable after. . . How is that a BIOS/Hardware issue?
(I know my system has been exactly the same since I installed the Sept 12 BIOS, and I would bet at least some of the others that are seeing this issue are in a similar situation)
BIOS/Hardware issues are systemic. . .meaning they impact every aspect of the system. . including everything that runs on that system. In my case, that’s just not happening. It’s just WoW Retail. . .and for me it just started this week, and I think others are having this exact experience.
So, If WoW Retail (and OW2 for those that play it) is/are the only thing(s) crashing on the user’s system then, just the opposite is likely true. . it points to a software issue not a hardware issue. On my system I have 21+ games installed including D4. . and WoW Retail is the only one doing this.
I don’t think that every Intel 13th gen CPU user is running things at their ragged edge, or constantly tweaking stuff. . I concede that some likely are and they may be doing things that are causing issues. . but I can speak for myself in saying that’s not the case here, and it’s likely the same at least for some of the others having this issue as well.
Is this issue fixable by a BIOS update? Maybe . . time will tell. . . but the circumstances really do point to they changed how they are doing something in their recent update that was missed by test cases and QA.
1 Like
There are also another couple elements to the mix: Windows updates and Nvidia/AMD making big changes in their drivers. Windows might have made some changes to things like their scheduler, memory mapping and WDDM related things. For instance, 23H2 is rolling out WDDM 3.2 and HLSL 6.8, along with some other features that might affect gaming. I think some of those features might already be in 22H2, since they are rolling this update out in a weird way; I’ll have to double check. This means they’ve likely made some internal changes to it all as well, even if you’re not using those feature levels.
I think the recent Nvidia 546.xx branch made some kind of big changes as well. You also have to remember that both Microsoft and Nvidia/AMD are constantly having to play whackamole with security holes. For example, a few months back, it was discovered that malware could resolve low quality images of what your browser is displaying through a side channel attack, if it was using hardware acceleration. Sometimes the ways they have to mitigate exploits can cause other random issues, but to them, it’s the lesser of two evils until they can figure out how to patch it correctly. Remember how hard the CPUs got hit during the spectre/meltdown exploit era? Up to 30% performance loss with the mitigation microcode updates and a bunch of random headaches from it all.
On a side note, an hour ago, I had a 0xc0000005 access violation error while playing CS2. I’ve never once had CS:GO/2 crash in all of the years I’ve played it and I’ve played it on and off since CS:GO was in beta(+2000h playtime).
1 Like
Yeup right on. . I totally agree the culprit could be any one of those as well. Some software interaction somewhere has changed enough that it has thrown things into a bad state and what ever it was. . it happened in the last 2-4 weeks. And it doesn’t agree with WoW Retail and Intel i9 13900s for sure. . .
Also on a side note. . I just used AWCC to overclock my CPU (and only my CPU) to it’s lowest possible overclock that AWCC will do. . .and it seems to have stablized WoW Retail, go figure. . . . which really makes me think you may be on to something with the scheduler. . I need to look at the November Win 11 updates to see if anything changed there with it. . there might be something that 10.2 doesn’t like and it’s manifesting on the 13900s because of the clock speed gap between the P and E cores, even though that’s “supposed” to be handled by the thread director. But I think it’s really odd that forcing OC, lowering the core ratio, and disabling speedstep/speed shift all seem to be palpable temporary work arounds for this issue in WoW Retail. Especially since it doesn’t seem to have been an issue just a few weeks ago. It really feels like a scheduler/timing issue.
2 Likes
Keep in mind that a lot of times, Microsoft doesn’t list a lot of changes they make to core security and kernel related things. Think of it like anticheat engines where they are always trying to stay a step or two ahead. There are also other angles like potential changes to HAGS that could be thrown into the mix as well. Oh and also, .NET 8 launched recently, but that might be unrelated to this issue (not sure if WoW uses it at all), but .NET is used a lot for thread workers in apps that use it, so if it’s not handled correctly, you can get errors due to race conditions or poor management of async tasks.
After having that random CS2 crash, I’m really starting to lean more toward this being a Windows/Driver issue. Maybe there are some things Blizzard could do to mitigate it on their end, but it would likely only be a bandaid fix and might lead to major losses in performance.
1 Like
Not saying you’re wrong, but how would this explain continual errors for people that have DDU’d and rolled back to previous Nvidia drivers?
If you have an Intel CPU you probably have Intel XTU though, and that’s easier and more convenient… unless you’re an Alienware and AWCC user apparently.
Microsoft changing something about Windows recently, would be my best guess. Naturally, companies like Nvidia and AMD are going to try to match those changes when they happen, or they can suffer from instability issues. It’s happened in the past plenty of times.
You play games through Windows and it handles everything like scheduling data to be executed and memory mapping requests(which then get virtualized by Windows). It’s the middleman between everything you want to do, unless you’re using something that’s ring-0 like kernel level anticheat softwares or kernel level drivers(most aren’t ring-0, but more like 1 or 2). WoW is user level, so everything it does goes through Windows. Any kinks in that Windows pipeline can cause instability issues in games like WoW, under the right circumstances.
Depends, but yeah it might be. Personally, if I’m going to mess with anything, I just use ThrottleStop because I can do things like FIVR profiles if I want them. I use a basic core/cache -offset undervolt, but sometimes I disable it to test things. I’ve had too many issues with the Intel XTU app in the past, so that’s why I just use TS.
1 Like
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.