AMD Radeon Crimson 16.1 Hotfix

Ohh CRSS.exe really?
Again is AMD fault 'cause they dont clean or update your computer!!! :wanker:

With old 15.12 I have some little flickering in e.g. Civ V,Mafia II,but with 16.1 all are good now.

 

this driver work good, fix all bugs for my games.. but some games dont run in max core clock like mortal kombat x run in 500 mhz but run at 60 fps. is normal ? or is a bug downclock the gpu clock?

 

If there's a framerate cap such as vsync (which "60" sounds like), probably working as intended in this case. Unless you feel some negative impact from it, glitches or corruption, stutter, or otherwise, then there's no reason to interfere with power-management and basically you should just let it do its thing.


--The downclock "bug" that people are talking about is when in games like Fallout 4, Borderlands, GTA-V, etc, their clock just drops to 300mhz sending the game in to single or low double digit FPS (for no apparent reason). Eg, where there's no CPU choke and the drivers are capable of keeping the card fed. (This seems to happen in specific games for the people that it does)

There are also still other cases where you may still opt for max speed, such as where you can feel frame-time inconsistency (this is especially important with freesync).


EDIT: Basically it's a try it and see situation. The key thing to keep in mind is that everything is a tradeoff, power-management included and the same goes with circumventing it (aka, more noise, more heat).

 

same here:

On Mon 11.01.2016 14:54:38 GMT your computer crashed
crash dump file: C:\WINDOWS\Minidump\011116-11531-01.dmp
This was probably caused by the following module: ntkrnlmp.exe (nt!KeBugCheckEx+0x0)
Bugcheck code: 0x109 (0xA3A01F59562076A0, 0xB3B72BDFA8A13039, 0x6, 0x18)
Error: CRITICAL_STRUCTURE_CORRUPTION
Bug check description: This indicates that the kernel has detected critical kernel code or data corruption.
This might be a case of memory corruption. More often memory corruption happens because of software errors in buggy drivers, not because of faulty RAM modules.
The crash took place in the Windows kernel. Possibly this problem is caused by another driver that cannot be identified at this time.

 

thanks, i already loosen my RAM timing, after that OTC BSOD, so far so good,

oh and besides if its RAM, it usually said IRQ less or not equal or memory controller management or something,

but that last BSOD, didn't mention anything related to RAM...

and 1 thing i still hate the most about OTC, is that i can't buy back the stocks that was bought by the AI :bang::bang::bang::bang:

OTC = Offworld Trading Company currently on Beta 9

 

AMD released another update for legacy gpus. Seems pretty weird for me cuz they have left support for older cards. I wonder if AMD want to contiune support by beta drivers.

http://support.amd.com/en-us/download/desktop/legacy?product=legacy3&os=Windows+10+-+64

Could anyone try this stuff?

 

Everythings fine on my HD6870. Just as fine as the flem modded ones.

 

Is this driver WHQL just like its GCN counterpart?

So far this 16.1 is being a good driver. It resolved my display corruption issue while idle (though AMD acknowledges it in a weird way, reverse psychology maybe). This issue affected all Fury's across the board.

This is called hotfix, I believe, because instead of new features this driver features a truckload of fixes AMD did last month.

I am not really satisfied by this weird support model. Often its competitor's "lawn is greener", literally, but progress made last year improving its overhead made me hope I can still stay with ATI/AMD (which I did, upgraded from 280X Toxic to a Tri-X R9 Fury).

 

According to the file that gets downloaded it's non whql so I didn't bother checking it. I think they will still support non gcn products for some time with their "beta" drivers.
(non-whql-radeon-crimson-16.1-ni-eg-64bit-with-dotnet45-win10-win8.1-win7-jan7.exe)

 

Interesting, I've still been having the flickering lines in some JRPG's and total corruption. Did not do very much testing without a voltage bump though, especially when idle and away from the machine on the desktop.


I'll have to play around a bit more and see if I see any differences with 16.1 in the number of occurrences. Glad that you've seen improvement, that's promising if some people are seeing an end to it.

 

Well, my Fury has been BIOS-modded (via some Python program) to fully enable all CU's, so I got somehow (playing with luck) an air cooled reference design "Fury X", running at 1040Mhz (Tri-X OC BIOS).

Maybe such mod "fixed" my card. So I'll run with the original BIOS (and "lose" the "free upgrade") for the next couple of days and report back if I get display corruption while idle.

What I have been testing so far is, if we still get "THREAD_STUCK_IN_DEVICE_DRIVER" bugcheck (BSOD), when we resume from S3 state (sleep).

1) It looks like it has been resolved, as long as you don't disable ULPS.
2) If you do disable, you'll get BSOD after resuming, and will reboot.
3) I don't know if Windows 10 updates interfered in such behavior (I don't know if AMD fixed the atikmdag.sys driver or Microsoft fixed DirectX kernel module).

I did very limited testing with my Toxic 280X, and it is behaving the same (Also ran 3Dmark, same performance if 1% better at best from 15.11). I forgot to test disabling ULPS, and then trying to resume from S3 state.

HD7970/280X owners (Hawaii/Tonga as well), especially if you got a custom/heavily overclocked card, please test resuming from S3 state, with ULPS enabled and disabled (You can toggle that via MSI Afterburner) and report back.

 

Modified inf for Wrestler/Sumo cards for Windows 10 http://forums.guru3d.com/showpost.php?p=5218156&postcount=38

 

Originally Posted by LIGuitar77 View Post
Whatever third party OC or other manipulation software you're using is causing it - or when you installed or reinstalled, etc...

No, I'm talking about anisotropic filtering settings.

 

I dont know if its the latest drivers, or a windows 10 update that happened the same day, but finally after 6 months of windows 10, pausing a video for longer than 30 seconds in mpc-hc, vlc, mp etc using crossfire actually resumes now (instead of image frozen and sound only resuming).

 

That is a known bug in madVR (http://forums.guru3d.com/showthread.php?p=5218252); the fact that it works for you now is most probably a side-effect of this driver breaking CF in games where it could be successfully forced before (as reported by some people in this thread).

 

fixed bug reappear?

it seem that the bug that start to appear at 15.11 crimson, but fixed at 15.12 crimson, is back at 16.1 crimson legacy for radeon 6600A windows 7 32bit, i am not sure about 16.1 /not legacy/ as i dun have the card to test it.

i think it is this one at release note.

[83111] System has higher idle power usage on Windows® 7 when running with Radeon Settings

http://support.amd.com/en-us/kb-articles/Pages/AMD-Radeon-Software-Crimson-Edition-15.12.aspx

UPDATE: it seem that enable or disable google chrome hardware acceleration, the timer resolution would still stuck at 1ms, it wouldnt goes up to 1ms, and drop back to 15.6ms when no needed.

 

I had the same here. Installed 16.1 and voilá all movies play correctly through mpc-hc and vlc. :banana:

 

Hi.

Don't know if anyone else has noticed, but the 16.1 driver has encreased my auto fan speeds. For my sake it was welcome because I think my card ran a bit too hot for my likings hitting 80-85c at load. Now it's stable around 75-78 after the 16.1 driver.

 

The glitch I had in Armored Warfare was fixed in this driver. It was similar to the compass glitch. I would get hit and the hp bar would go all wonky for a while so I had no clue how much I had lost or if any. Good stuff. All in all best driver I've had from AMD

 

Do these drivers fix the stuttering due to clocks downclocking? Or should I be using ClockBlocker to fix that?