rigaya/QSVEnc

mfx ai superres not working in Intel dGPU(A380)

Closed this issue · 2 comments

I tried it with a 60-second test video, but I'm getting the error log below.

==================================

E:\C\Downloads\QSVEncC_7.71_x64>QSVEncC64.exe -i test.mp4 -o result.mp4 --output-res 3840x2160 --vpp-resize mfx_ai_superres

result.mp4

QSVEncC (x64) 7.71 (r3471) by rigaya, Oct 27 2024 12:02:14 (VC 1941/Win)
OS Windows 11 x64 (26100) [UTF-8]
CPU Info AMD Ryzen 5 9600X 6-Core Processor [5.57GHz] (6C/12T)
GPU Info Intel Arc A380 Graphics (128EU) 300-2450MHz (32.0.101.6129)
Media SDK QuickSyncVideo (hardware encoder) FF, 1st GPU(d), API v2.13
Async Depth 3 frames
Hyper Mode off
Buffer Memory d3d11, 33 work buffer
Input Info avqsv: H.264/AVC, 1920x1080, 30/1 fps
VPP AI SuperRes, mode 1
Resize 1920x1080 -> 3840x2160
AVSync auto
Output H.264/AVC(yuv420) High @ Level 5.1
3840x2160p 1:1 30.000fps (30/1fps)
avwriter: h264 => mp4
Target usage 4 - balanced
Encode Mode ICQ (Intelligent Const. Quality)
ICQ Quality 23
QP Limit min: none, max: none
Ref frames 3 frames
Bframes 3 frames, B-pyramid: on
Max GOP Length 300 frames
Ext. Features PerMBRC RepeatPPS
Break in task MFXVPP: gpu hang..

encoded 0 frames, 0.00 fps, 0.00 kbps, 0.00 MB
encode time 0:01:00, CPULoad: 0.1%

QSVEncC.exe finished with error!

=======================================

Yes, it seems like scaling more than 1080p causes the GPU hang error.

It shall work fine with SD → HD, but not with HD→4K. Please use other scaler for that.

I'll close the issue as the topic has been answered.