I have a "Sapphire HD-7950 Vapor-X 3GB with OC boost" GPU card in my cMP 5,1 (MacOS 10.13.6 Hgh Sierra) which I bought used and am trying to figure out as it's not behaving as expected when it comes to attaching two displays to the two DVI connectors.
Its behaviour differs when in Windows 7 compared to MacOS, and also the various BIOS (ROM) files I've tried out. It's quite confusing 😓
As the card has had two previous owners I'm not sure of its history; whether it came with the factory-installed BIOS or if it's been modified. I've been able to get hold of what is believed to be the original factory-installed BIOS files from TechPowerUp.com, but if I reflash the card with it I only get one working output (also in Windows).
On the other hand, the BIOS that came with the card works with both displays in Windows but works strangely in MacOS where the "Displays" system preference tells me I have two displays attached (both with the correct model numbers), but one of them remains black.
As I now have several BIOS files (850MHz and 950MHz (OC boost) versions of the original owner's, the downloaded factory-installed ones and Mac -EFI bootscreen patched ones of all the above I've created from info in this forum) I see different behaviours and am in the midst of taking notes while comparing the behaviour in Mac and Windows.
I have read that some people experience that certain output connectors no longer work after having Mac EFI-patched and reflashed their current BIOS, but what puzzles me is that I get differing results (black screen/screen working) when using non-Mac flashed BIOSes when booting into Windows or MacOS.
I'm thinking there's something I've missed or don't understand.
Could it be that with Windows, some information is stored within Windows or its display driver setting? Or, does the PRAM memory store some information which is only relevant to Windows regarding display outputs which is disregarded by MacOS?
I ask the latter because the Windows app GPU-Z (which shows all sorts of info about an installed GPU card) tells me the current BIOS is UEFI enabled. Doesn't that have something to do with the computer's BIOS, and would this be the cMP's PRAM memory?
Its behaviour differs when in Windows 7 compared to MacOS, and also the various BIOS (ROM) files I've tried out. It's quite confusing 😓
As the card has had two previous owners I'm not sure of its history; whether it came with the factory-installed BIOS or if it's been modified. I've been able to get hold of what is believed to be the original factory-installed BIOS files from TechPowerUp.com, but if I reflash the card with it I only get one working output (also in Windows).
On the other hand, the BIOS that came with the card works with both displays in Windows but works strangely in MacOS where the "Displays" system preference tells me I have two displays attached (both with the correct model numbers), but one of them remains black.
As I now have several BIOS files (850MHz and 950MHz (OC boost) versions of the original owner's, the downloaded factory-installed ones and Mac -EFI bootscreen patched ones of all the above I've created from info in this forum) I see different behaviours and am in the midst of taking notes while comparing the behaviour in Mac and Windows.
I have read that some people experience that certain output connectors no longer work after having Mac EFI-patched and reflashed their current BIOS, but what puzzles me is that I get differing results (black screen/screen working) when using non-Mac flashed BIOSes when booting into Windows or MacOS.
I'm thinking there's something I've missed or don't understand.
Could it be that with Windows, some information is stored within Windows or its display driver setting? Or, does the PRAM memory store some information which is only relevant to Windows regarding display outputs which is disregarded by MacOS?
I ask the latter because the Windows app GPU-Z (which shows all sorts of info about an installed GPU card) tells me the current BIOS is UEFI enabled. Doesn't that have something to do with the computer's BIOS, and would this be the cMP's PRAM memory?
Last edited: