That's ok but any benefits (faster speeds, better stability) for mining?
Exactly the opposite - 5-10% LOWER hashrate vs. 16.9.2 (or 15.12 on pre-RX cards).
IMO miners should just IGNORE this bloated slow INTRUSIVE piece of junk.
Topic: Latest amd driver 16.12.1 / (Code 43) for AMD cards under device manager (Read 3048 times)
December 12, 2016, 03:06:53 AM
Exactly the opposite - 5-10% LOWER hashrate vs. 16.9.2 (or 15.12 on pre-RX cards).
IMO miners should just IGNORE this bloated slow INTRUSIVE piece of junk.
December 11, 2016, 09:04:06 AM
I think this new driver improve power use, so are the card consumes less W in mining?
From my limited observations with the new driver so far it doesn't look like there was any change in power draw (up or down) while mining versus the 16.11.5 version. For me I am staying with the older driver for now as it is too much of a hassle to update all my rigs for little to no benefit.
December 10, 2016, 02:03:57 PM
I think this new driver improve power use, so are the card consumes less W in mining?
to some extent it did improve power use not enough to matter much and increased my speeds on some cards mainly my 400 series cards not much but some not enough that it really matters were it's a big deal.
I guess you would call it more what you like then anything and it does seem more stable to me but others have said they had nothing but issues.
December 10, 2016, 12:18:45 PM
I think this new driver improve power use, so are the card consumes less W in mining?
December 10, 2016, 12:00:37 PM
That's ok but any benefits (faster speeds, better stability) for mining?
December 10, 2016, 07:27:56 AM
totally up to you i just gave you other way to use it if you want to use it without losing anything.
i like the driver, some don't .
i like the driver, some don't .
December 10, 2016, 06:26:26 AM
or do this
if you have a modded bios AMD turned on drivers signing so you can do one of two things use a modded version of the Drivers or sign your own bios which is possible but harder to do over patching the atikmdag.sys kernel .
you can by pass it like this
http://www.overclock.net/t/1604567/polaris-bios-editing-rx480-rx470-rx460/1050#post_25699216
and it does work .
I'm using The new AMD driver with a modded bios after i patch the driver, i Like the new driver and using it with no issue at all other then the driver signing junk .
if you have a modded bios AMD turned on drivers signing so you can do one of two things use a modded version of the Drivers or sign your own bios which is possible but harder to do over patching the atikmdag.sys kernel .
you can by pass it like this
http://www.overclock.net/t/1604567/polaris-bios-editing-rx480-rx470-rx460/1050#post_25699216
and it does work .
I'm using The new AMD driver with a modded bios after i patch the driver, i Like the new driver and using it with no issue at all other then the driver signing junk .
Why would i do all that patching stuff ? Is it worth it ? Those drivers are better for mining or ?
December 10, 2016, 06:12:32 AM
I reverted my VBIOS back to stock settings on my test rig and managed to get the 16.12.1 drivers to install ok, but the bad news is I seem to have lost about 10-15 sols using the same settings as before. Using a MSI RX470 I am down from 215 sols (1250/1750) under 16.11.5 to only 200-205 sols with the 16.12.1 driver. Both tests were when using the stock bios for a fair comparison and Claymore v9.1. I am going back to the older driver for now, maybe in another release or two AMD will optimize the drivers some more and get rid of the annoying integrity check.
December 10, 2016, 06:09:05 AM
or do this
if you have a modded bios AMD turned on drivers signing so you can do one of two things use a modded version of the Drivers or sign your own bios which is possible but harder to do over patching the atikmdag.sys kernel .
you can by pass it like this
http://www.overclock.net/t/1604567/polaris-bios-editing-rx480-rx470-rx460/1050#post_25699216
and it does work .
I'm using The new AMD driver with a modded bios after i patch the driver, i Like the new driver and using it with no issue at all other then the driver signing junk .
if you have a modded bios AMD turned on drivers signing so you can do one of two things use a modded version of the Drivers or sign your own bios which is possible but harder to do over patching the atikmdag.sys kernel .
you can by pass it like this
http://www.overclock.net/t/1604567/polaris-bios-editing-rx480-rx470-rx460/1050#post_25699216
and it does work .
I'm using The new AMD driver with a modded bios after i patch the driver, i Like the new driver and using it with no issue at all other then the driver signing junk .
December 10, 2016, 05:44:57 AM
This mode does not work on the latest drivers. Only 5 cards
December 10, 2016, 05:37:03 AM
Hello ,
Try this little program, it is to moderate the driver,
I had this code 43 when I added my 5th card, could not be detected, and with this mod installed code 43 disappeared.
https://lbr.id.lv/6xgpu_mod/
Try this little program, it is to moderate the driver,
I had this code 43 when I added my 5th card, could not be detected, and with this mod installed code 43 disappeared.
https://lbr.id.lv/6xgpu_mod/
December 10, 2016, 05:16:56 AM
16.12.1-5December is trash to mine with even when it does work - gives noticeably lower hashrate vs 15.12 on my pre-RX cards (I don't HAVE any RX-series cards yet, first once should be arriving early next week).
December 10, 2016, 03:56:09 AM
I got this error too when trying the new ReLive drivers on a 3 GPU test rig that I use to try out new things such as this. The interesting thing was I was able to get two of the cards to work, but never all three at the same time. I was using GPUs with a modified BIOS, but did not try the disable driver integrity step. I figure I will wait until AMD either removes the check in a later release or someone else reports enough of a performance improvement to make it worthwhile. Went back to 16.11.5 which I think was the last working version (for mining) before the 16.2.x ReLive versions.
December 10, 2016, 02:45:21 AM
the interesting part is , on both rig that card worked fine which one was connected to a dummy plug or a monitor . all another cards got a CODE 43 error under device manager ... all cards have a custom modded bios by me ..
December 10, 2016, 02:38:03 AM
If you have modified roms you will have to use Eliovp's instructions for disabling driver integrity check.
There are other ways too. I believe this can be done in power shell and also by changing a windows setting in Windows 10.
Or.... you will have to roll back to a version like 16.9.2.
There are other ways too. I believe this can be done in power shell and also by changing a windows setting in Windows 10.
Or.... you will have to roll back to a version like 16.9.2.
December 10, 2016, 02:34:14 AM
Im not sure its just me , or anyone else have this problem , few minutes ago installed the latest amd driver (16.12.1) on 2 of my rigs ( 1 with 4 rx 480 , another with 6 rx 480 ) and on both build i got error 43 under device manager . the card on both rig with a monitor or dummy plug plugged in is without that error , all another cards got a error " CODE 43 "
UPDATE : rolled back the driver to 16.9.1 all cards working fine , no "CODE 43" error ....
UPDATE : rolled back the driver to 16.9.1 all cards working fine , no "CODE 43" error ....
Jump to: