Radeon Software Crimson Edition 16.9.1 Download & Discussion

If i am about to buy something it will be a RX480 8GB not Fury unfortunately.

You are right tho, we need an upgrade but like i said before, i dont play some super demanding games (only Doom but that game is flying on Vulcan) so i can wait little more.

And PrMinisterGR that quadrobake 7970 always make me laugh. Its a miracle card is still functioning, lol. Mine is burning at 1.3v for last 3 years and it refuses to die.

 

Uff OK

 

Compared to 16.8.3 perhaps, but actually I had to use ClockBlocker to squeeze out that "gain". Some driver from Jun shows a different picture though.

http://www.3dmark.com/compare/aot/164731/aot/164729/aot/141873#

 

Undying, what's your highest OC without touching the voltage? And what's your card's ASIC rating?

 

I have a problem since the Anniversary Update.

Every driver I tried restarts my pc when playing games like DotA 2 and GTA 5.

GPU - Sapphire 7870 GHz Edition

P.S. I never encountered this sh*t before the update. TSK!

Please advise.

 

Asic is 65.4, so not that good. I didnt even tried to overclock without touching the voltage. I always ramp up the voltage and try to reach my maximum. In this case its 1165mhz. Anything higher gives some wierd artifacts even at 1.3v.

 

My highest OC is 1120/1600 without touching voltages. ASIC 56.2%
Increase voltage don't make any sense with my GPU cooling system.
This driver fixed idle clocks bug for me after update Windows to AU.
WDDM 2.1 here too.

 

If your machine outright reboots without any BSOD, it is most likely because your power supply is failing and all this happening right after the Anniversary update is probably nothing more than a coincidence. If the hardware info in your profile is up-to-date, my money is on the power supply since you are not using such a high quality unit. If this is the case, you should stop using the machine right away and replace your PSU before it dies and potentially kills the rest of your machine.

If your machine BSODs and then reboots, however, that is a whole different scenario. You should check the memory dumps with WhoCrashed or BlueScreenView.

 

My profile is a little bit outdated. Sorry. That PSU already died on me. Let me update my profile and please check.

I recently had a clean install and updated the drivers to this release and played GTA 5 and it restarted again without any BSOD.

Gameplay is around 30 minutes.

 

That seems in line with what I posted here http://forums.guru3d.com/showthread.php?t=409380

Yes I had that issue before only on the API Overhead, and to be honest I don't know what I did to resolve it. What I know is that now I don't run MSI Afterburner with any 3DMark test and I disable Kaspersky too.

You can also try and disable monitoring in 3DMark.

 

Bug doesn't seem to happen on my system. Clocks always go back to 150/300.

 

Which software are you using for increasing voltage? Msi afterburner has everything locked for me, even with settings "customization".

 

The fact that the problem did not go away with a clean install rules out software-related causes. Random reboots without BSOD almost always points to failure in power delivery components, nearly always the PSU and maybe sometimes the motherboard. Corsair's VS and CX PSUs are the bottom of their line-up and despite bearing Corsair's name, they are not actually such good-quality units.

If you could get a spare PSU from somewhere to test, that should help you in determining the case. If not, assuming that your PSU is still under warranty, I recommend RMAing it. Needless to say, you should avoid pushing your machine until you determine the cause of the failure. You should also back up your important data externally if you haven't; lost hardware can be bought back with money, lost data... is lost.

This only seems to happen on GCN v1.0 (7xxx & R9 280/X).

 

You have a 390 right? The ones that we seem to get it are people with Tahiti/Pitcairn.

 

@THEAST
Thanks for the info.

@PrMinisterGR
Yup, 390 here. Thanks for clarifying.

 

i dont have the clock bug either, screen just woke up and its 300/150. with hardware acceleration on in firefox it runs gpu at 950/1550,stays like that in games. dont know if this happened in other drivers since i had it off for a long time.

 

Looks like WDDM 2.1 drivers (both NV and AMD) breaks some d3d9 games (eg: Fallout New Vegas).

atiu9pag.dll call of Direct3D 9 OpenAdapter() caues a 0xc0000005 exception (access violation).

 

OT: It's funny how ASICs go on Tahiti. On mine without touching voltage I can't even get up more than ~1025MHz without artifacting after some time of gameplay. The same for memory speeds, because without increasing core voltage, instead of ~1675MHz (this card uses 1.6V, cooled chips but connected to the VRM plate...), even 1500MHz is recipe for failure when already nearing the gpu clock limit at stock voltage. That's on a 82.x% ASIC card with 1.050V default (about 1.026V on load). 1.15V yields about 1175/1675MHz stable last time I've tested, but power consumption skyrockets and, other than in benchmarks, I'm too CPU limited to make it worth it.


On the clocks problem, I just find it weird that you got it solved after AU. On my Tahiti and Pitcairn AU systems it seems the other way around with recent drivers, as both are showing the same problem.

 

I also have the idle bug on my 280x but I also noticed that when the idle bug was active, my card wouldn't go to it's boost clock in games, it would only go to the regular non-boost clock. I restarted the machine and it used the boost clock in games again. I haven't tried to reproduce it again to see if it is because of the idle bug.

 

I've been having this bug for quite a while now and 3d clocks are working fine thats the reason i stopped caring.

This is the solution. Notice how UVD clocks are the same as performance :

http://forums.guru3d.com/showpost.php?p=5329225&postcount=7