AMD Radeon Crimson 17.7.1 Download & Discussion

Installed over 17.6.2, everything working ok. CRU settings didn't get obliterated either. Just had to reapply my 1620p VSR registry setting.

 

5-7% worse performance\less fps at maximum GPU load in D3D11 apps compared to 17.5.2, just like with 17.6.1(2)

 

Got the same error. For me cause was disconnected HDD (old 750GB SG used for temp downloads) where I ran all previous AMD driver auto extracts from. Caught this by trying to manually install ccc-next64.msi from c:\amd\..driver version..\Packages\Apps\CN\CNext\cnext64. That gave error abour E:\ not accessible. Fixed by connecting USB stick and assigning E:\ for it.
Try to manually install cnext64.msi to see what errors it comes up with.

 

Tried what you mentioned and I got this error:

Internal Error 2203. C:\Windows\Installer\inprogressinstallinfo.ipi, -2147287037

 

Finally, all 17.6.x was great ones and now this one 17.7.x is Great also = :banana:
Better performance + OC across the board !

565MHz HBM with tMOD 400MHz (300Mhz is default HBM internal Timing Strap)
in CryEngine

 

- I had the same problem..
- download the new driver from AMD's site and save it..
- uninstall the existing driver
- on reboot go to downloads and run the file you got from AMD

- seems to work better..

 

What program do you use to check this?

 

I get the same error when I uninstall AMD. The problem might be related to the MSI Installer. I get errors in my Event Log (10005). As I said the error is occurring during the install of the AMD settings. All the other components installed fine such as HDMI and the driver itself.

 

Solved

I fixed my problem. It might have been fixed by a couple of things I did.

1. A few weeks ago I set my Windows\Temp Folder to point to the one in my Appdata\Local folder by changing the environmental variables (TMP, TEMP). I changed the Windows Temp folder back to it's original settings.

2. I gave Full Control to Everyone in the Windows\Installer folder as well as the two Temp folders from above.

 

Any way to reproduce this? I don't seem to have this issue on my 7950. Do you guys have a custom bios or use OC software?

 

For me it can be reproduced 100% reliably while making Skype video calls. As soon as I start such call, the GPU clocks go to UVD clocks as expected, but never come back down even if I end the call and close Skype. After this, my 3D clocks will also be reduced from 1000,1250 to 900,1250. No custom BIOS or OCing here; I did downclock my cards before to prevent extra heat, but recent versions of AMD's driver also break compatibility with MSI Afterburner on my system (another feature that worked but was broken, like the UVD bug), so I am running default clocks now. AMD seems to be very keen on breaking stuff on these old cards.

Others have mentioned similar behavior with GPU-accelerated in-browser video playback, or GPU-accelerated video rendering using MPC (with VMR/EVR; madVR is unaffected) and other players.

 

Best is to have:

Administrator + Everyone + Users (All Permissions)

Windows
Program Files
Program Files 32
Users
ProgramData

Everything is working Great (on any WinX build)

 

I think I would rather have clocks that are stuck high then clocks that stay too low when using apps. and then getting TDR because of it..:bang:

 

No problems so far with this driver...everything seems stable for me.

 

From a security standpoint, your setup is a horror story, sorry to say.

 

hahah

 

He should be thankful. I'm not sure if I've ever seen worst tip. He has just committed... a felony! Anyway I really don't have any idea why that... tip could resolve any problem.

 

lol :bang:

 

Basically, by giving "Everybody" account full access to all program files, and the Windows OS directory, you pretty much leave the OS wide open to attack by any process, or telnet access, or any other backdoor trying to access sensitive areas in Windows.

If OnnA wants to complete the security stripping, he should also add "Network" to that setup with full rights.