MSi Afterburner keep crashing after i updated to windows 10 version 1809 it automatically closes itself after few min without any error message i have tried latest beta as well but no luck
Closes? Is notice in logfile? Some game anti-cheat systems kill AB. Maybe some anti-cheat went crazy on 1809. Apparently, if it can't be replicated, you have to provide meaningful source of information leading to root cause. So, check system logs, AB logs, ...
It is not crashing. New version of BattlEye is silently forcibly closing it on AMD systems without notifying user in any form.
Hi! I noticed that if I use the MSI afterburner with the latest AMD drivers, on overclocking I'm not stable. But if I try the old drivers (16.11.2), before Wattman is implemented I can have a stable overclock. Has anyone of you solved? How can I solve? Does anyone have my same problem?
You should consider other cause. New driver utilizes your GPU better => more load/heat => higher v-drop => less stable. Those things happen all the time, especially if your GPU aged a bit and requires higher voltage on same OC. Or maybe no voltage is enough on given clock and you'll have to downclock a bit.
Not sure if has been posted before, but the synchronized dual fans for the MSI GeForce RTX 2080 Ti Gaming X TRIO don’t work as they should in manual mode. When setting a custom fan curve or manual speed, fan #1 shows 350-500 RPM higher speed than fan #2, bigger difference the higher you set the fan. At 100% fan speed you get some 3400 RPM vs 2400 RPM. Faster fan is that small one near the front of the card. When left in the standard auto-mode, both fans spin at proximately the same RPM, all the time. Usually in the range 1350-1650 RPM, depending on temperature. Maybe you should add a Fan 2 tab with its own separate graph, and just leave the synchronized mode as an option.
You’re misunderstanding things. Who said that you should get exactly the same RPM on different fans? Fan duty cycle (speed in %) must be the same, but the same % is not supposed to be mapped to exactly the same RPM.
You are correct. On auto, fan speeds are something like 46% and 59%, for practically identical 1500 RPM on both tachometers. But wouldn’t it be better if we could replicate this percentage offset when setting a custom fan curve? That faster fan becomes noisier much sooner than the other two and is not ideal. If the MSI vbios thinks both fan groups should have the same RPM during gaming, then perhaps it makes sense to duplicate that when manually adjusting for lower temps. I would imagine other cards might need a different offset, or none at all. But at least there should be some way to tweak that when its needed. BTW, that OC scanner feature is pretty much spot on with the frequency curve. Great work. Thanks!
That doesn’t work that way at all, there are no and there won’t be any offsets or attempts to sync RPM and there is no any default “RPM sync” mode like you’re assuming. Sync will always result in syncing fan duty cycle only and I have nothing to add to it.
I've noticed that with the latest AMD Drivers (18.10.1 & 18.10.2) the GPU Usage graph in Afterburner is grayed out , while in Wattman and task manager the GPU Usage is displayed just fine. The issue doesn't exist with older driver versions such as 18.9.3. I was wondering if anyone else has this issue , and if there is any workaround for it ? RX 480 8GB MSI Afterburner 4.6.0 Beta 9
Randomly MSI Afterburner stopped working, here is error popup. I tried reinstalling 4.50 and beta but it still wont wor k.
hi there, I'm about to replace my current EVGA GTX 970 video card by a new ZOTAC RTX 2080 AMP 8GB in a couple days. I'm currently using Afterburner 4.4.2 and RTSS v7.0.2 which doesn't support the Turing video cards. Should I uninstall current version before installing Afterburner v4.6.0 beta 9 + RTSS v7.2.0 final for the new video card ? I never do overclocking or use custom fan profiles, everything is set to auto. I'm using Windows 7 SP1 64-bits but plan to do a dual boot with Win10 build 1809 later. Thank you.
It's recommended but most often not at all needed. Give it a go, if you see an issue you can always uninstall or clear the RTSS cache.
It's a feature using extensions in NVAPI (Used to get data from the GPU among many other things.) and certain code for Turing that's under NDA at least for now though it remains to be seen if NVIDIA might open it up more or if it's going to be closed, probably a pretty complex piece of work and not something that they might want to allow modifications too for the sake of GPU reliability such as being able to change the limits it has. It can't be directly implemented on AMD and while I am no expert on the subject matter I do not believe AMD's ADL API has the same flexibility for overclocking which is one part of why you need low level hardware access for some of the GPU data in Afterburner if I'm not messing this up completely. (I do know NVIDIA has some pretty interesting things in the NVIDIA API which also goes beyond just getting data from the GPU itself.) A simpler overclock suite might be feasible, AMD had a very basic 3D demo in their early drivers for example but it wouldn't be anywhere near the same as NVIDIA's AI suite of self diagnosing and testing for stability found on Turing. (From that to Overdrive to Wattman and it hasn't exactly been smooth for AMD plus underlying changes in the driver occasionally breaking support with third party utilities.) EDIT: It's both a hardware and software limitation and it's a feature found in NVIDIA's Turing GPU's that can be called from Afterburner but it's under NVIDIA's control and it's their own code. (There's better and more detailed info in the Rivatuner forum here and the beta topic for this version detailing how this is set up and how it works.) EDIT: Thinking about it since it's pretty closed I wonder if it might not even go via NVAPI at all and use something entirely different, might be more low level entirely than what I'm thinking on how it's working. Yes I think that's making a bit more sense at least for the AI overclock feature itself and it's scanner and test suite. Should just have said that it's a closed-source private feature NVIDIA implemented for Turing which AMD can't use.
Would have been nice if MSI had actually said that on their site or if had of actually been mentioned on the download page although ATI used to have a OC scanner I wonder why they can't make use of it for AMD cards
