Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Since above 4G decoding and ReBar are a thing now... #56

Open
gustarballs1983 opened this issue Mar 28, 2024 · 0 comments
Open

Since above 4G decoding and ReBar are a thing now... #56

gustarballs1983 opened this issue Mar 28, 2024 · 0 comments

Comments

@gustarballs1983
Copy link

...will setting higher ReBar limit will increase GPU-Ramdrive Performance..

Since most folks misunderstand how ReBar works, and it is common for people to think that it enables access to GPU vRAM above 256MB, its time to straighten up those ejaculates....

All that ReBar does is increase the maximum size of a single packet in packet transmission between GPU and the system's DRAM, single packet used to be limited to 256MB, but now it can be set as big as imagination allows ( now sky is the new packet size limit).

Hence why i think it should drammatically increase GPU-Ramdisk performance... Am I right?

Unfortuneatly I don't yet have a system with Above 4G decoding & ReBar enabled. I'm working on it, The only system that i have with UEFI, doesn't have the options to enable those... yet..

I'm in the process of adding support for it to the UefiBios, in form of commonly known UefiBios ReBar mod, and Above 4G Decoding is as old as UEFI itself, it's just not visible to toggle and disabled by default, so agin bios needs to be hacked in order to enable it in my case. Other than that I may test GPU-Ramdisk with ReBar enabled with AMD's RX590-GME, and possibly with Intel's Arc A750, if GPU-Ramdisk devs add support for Intel dedicated cards..

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant