A few have commented here and on Youtube and Reddit stating it would be nice to see a Terascale Vulkan driver. There are enough HD 5000 and HD 6000 cards in the wild still, and are very cheap. These would make a great entry level gaming card for someone (or a budget emulation station). (Historically, the purchase of ATI nearly resulted in AMD's financial ruin: there were not enough resources internally to do Vulkan on GCN and legacy platforms. But now it seems a broader effort could be achieved)
As Steam pushes gaming on Linux, getting older cards up and running for gaming (and getting the AMD brand out there as a long-lived gaming platform) seems like it would be beneficial for its "halo effect" for newer product purchases. Many budget and older games could still run fine... and... it would be fun to see benchmarks of course.
This is a DX11-era card, and plenty of modern games still use DX11 or older.
It has been stated that a Vulkan driver for these devices is technically feasible. So: can anyone outline a list of roadblocks, other than financial (no sponsoring organization)?
(I understand that "free" software development has a large cost: especially for highly talented developers that would need to work on this. I'm looking only for the technical limitations / what workarounds would be needed that make it so... cost-prohibitive to implement.)
As Steam pushes gaming on Linux, getting older cards up and running for gaming (and getting the AMD brand out there as a long-lived gaming platform) seems like it would be beneficial for its "halo effect" for newer product purchases. Many budget and older games could still run fine... and... it would be fun to see benchmarks of course.
This is a DX11-era card, and plenty of modern games still use DX11 or older.
It has been stated that a Vulkan driver for these devices is technically feasible. So: can anyone outline a list of roadblocks, other than financial (no sponsoring organization)?
(I understand that "free" software development has a large cost: especially for highly talented developers that would need to work on this. I'm looking only for the technical limitations / what workarounds would be needed that make it so... cost-prohibitive to implement.)
Comment