yuzu/yuzu-emu.github.io
1
0
Fork 0
mirror of https://github.com/yuzu-emu/yuzu-emu.github.io.git synced 2025-07-30 05:32:06 +00:00
Code Issues Packages Projects Releases Wiki Activity

fish x2

Browse source
This commit is contained in:
Matías Locatti 2021-06-07 03:24:20 -03:00 committed by liushuyu
parent 35f6be8285
commit 2bc7023d9b
1 changed files with 2 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
site/content/entry/yuzu-progress-report-may-2021/index.md
View file

@ -58,9 +58,9 @@ Thanks to user reports, it has been brought to our attention that our FPS counte
The readings were accurate in the past, when the GPU thread was synchronous with the other services. The readings were accurate in the past, when the GPU thread was synchronous with the other services.
Since multicore support was implemented, this is no longer the case. Since multicore support was implemented, this is no longer the case.
The solution epicboy wrote is simple, [display the average framerate between the previous and latest update](https://github.com/yuzu-emu/yuzu/pull/6317), The solution epicboy came up with is, instead of letting the nvflinger service handle the counter, [let the graphics API inform the counter after it finishes drawing the frame](https://github.com/yuzu-emu/yuzu/pull/6317),
this way the value is far more accurate and stable on performance oscillations. this way the value is far more accurate and stable on performance oscillations.
Additionally, the update frequency was changed from 2000ms to 500, giving more updated info on the reading. Additionally, the update frequency of the counter was changed from 2000 ms to 500 ms, giving more information on the reading.
A big one for Vulkan users waiting for `Project Reaper`to leave the development board. A big one for Vulkan users waiting for `Project Reaper`to leave the development board.
The previous Vulkan `memory allocator` used local memory as top priority (that’s the physical VRAM on your dedicated GPU), The previous Vulkan `memory allocator` used local memory as top priority (that’s the physical VRAM on your dedicated GPU),