I recently bought an ASUS G15 AMD Advantage Edition with the R9 5900hx and 6800M. There have been some pretty intense driver issues when playing demanding games such as Assassin's Creed: Valhalla. The problem is difficult to replicate because it is so random, and it seems the worst when playing Valhalla. Without frame limiting and power reduction through the use of Armory Crate, I can usually only play in the realm of 5 - 15 minutes before I get a crash to desktop and an AMD message saying that the driver has timed out. In event viewer, the corresponding error of "display driver amdwddmg stopped responding and has successfully recovered." When I apply some sort of limiting factor to the GPU to not work as hard, the game will run for longer, in the realm of 35 minutes to an hour, but will still crash.
It is my theory after two weeks of troubleshooting, failed attempts, DDU safe mode reinstalls of both (there are only two driver versions available for the 6800M) drivers, Windows reinstalls, game file integrity checking, supporting software reinstalls, and mitigation of conflicting software, it is my theory that AMD's or ASUS's clock speed profile of the card is to blame.
The 6800M, according to GPU-Z, is rated at a 1981mhz base clock, 2116mhz game clock, and 2390mhz boost clock. Well, the card in this laptop regularly hits 2575mhz and sits there. While I appreciate either ASUS's or AMD's (I'm not sure who is to blame for the aggressive overclock profile on this card) attempt to squeeze the most out of their machine for their customer, it has created an unstable gaming situation. Normally, this would be easily amended by a third-party program or even AMD's software. However, AMD has locked all card control on their mobile units, forcing the user to comply with whatever speed profile has been deemed acceptable by the manufacturer - in this case, the unstable one created by either ASUS or AMD.
I believe this to be the issue because of the mitigation that controlling the framerate allows does to the driver timeouts. Controlling the framerate essentially forces the GPU to work less hard and artificially lowers the clock speed. However, it is a clumsy and inefficient way to do so and leaves gaps for errors, especially when in graphically demanding parts of any game. Not only that, but it does not affect the voltage at which the card runs. Armory Crate allows for controlling the combined power draw of both the CPU and GPU package, but once again, it is a clumsy way to force the clock speed down and does not directly address the problem. The same goes for Radeon Chill and other FPS limiting software.
The only ideal solution would be to lower the clock speed down to its boost clock or lower for a stable gaming experience across the board. With the 6800m, this is simply not possible due to AMD's staunch lockdown of its mobile GPUs. Infuriatingly, I can adjust the clock speed on the integrated graphics, but not the 6800M. Why AMD decided to allow one and not the other that matters is beyond me.
AMD. Please. Please either allow your mobile GPU's clock speed to be controlled by your customers to create a more stable experience. Or at least, in the alternative, allow your customers to flash their mobile GPU bios to clock speed profiles more reasonable or that don't go past the boost clock. You've given some of your customers cards that literally break themselves from going too fast and there's nothing your customers can do about it.
Hello i want to join this,pardon my english im not native,
i got the same laptop asus advantage 5900hx and 6800M, im having similar issues but the most anoying ones are the heavy fps drops in e-sports titles like league of legends, i play above 200fps but in teamfights my fps drops to 0 and is not msi afterburner or anything looks like a problem with low power states in the gpu and there seems no way to change it,there is awful performance in path of exile 100-120fps in lower settings and no load limit always on, also ive noticed huge loss of performance with videos playing in the background , there is also an issue with the voltages of the cpu, when switching from really demanding cpu games to idle is really agresive the change in power and voltages and i am still with high temperature so i get cpu errors like the ones i get from a really hard undervolt maybe because of temperature increasing the resistance and the laptop freezes and reboot. There should be atleast a way for us consumers to adress this kind of issues , a better way than to flash the gpu ...
Hi! Do you still have that problem? Did you manage to solve it? I have the same laptop and the same problem but with Apex, I already updated drivers, windows 10 and nothing, I also did some benchmarks thinking that the CPU or GPU was wrong and it turns out that everything is fine.