Try the same command but remove overlay filter and check. Marcin
W dniu 28.02.2018 o 14:44, Garri Djavadyan pisze:
Hello FFmpeg community, I faced a problem with NVDEC/NVENC resources underutilization while running one ffmpeg instance. We use ffmpeg to convert various format videos to MP4(h264/aac) and applying logo overlay. Hardware decoding and encoding process is performed by NVDEC and NVENC chips. For example, our cmdline is: /usr/local/ffmpeg-dev/bin/ffmpeg -y -c:v mpeg4_cuvid \ -i input.avi -i logo.png \ -filter_complex [0:v:0][1:v:0]overlay=10:10[out1] \ -map [out1] -map 0:a:0 -map_metadata -1 -map_chapters -1 \ -c:v h264_nvenc -b:v 1024k -r 25 \ -c:a libfdk_aac -b:a 128k \ -movflags faststart out.mp4 The overal transcoding process is greatly accelerated. But I see, that NVDEC/NVENC cycles are not fully utilized. For example: # nvidia-smi dmon # gpu pwr temp sm mem enc dec mclk pclk # Idx W C % % % % MHz MHz 0 32 50 11 3 22 16 3802 1632 0 32 50 11 3 23 14 3802 1632 0 32 50 12 3 22 15 3802 1632 0 33 51 13 3 18 15 3802 1632 0 32 51 12 3 17 11 3802 1632 0 32 51 10 2 20 13 3802 1632 I tried to find a bottleneck, but all system resources are OK. For example, CPU (top output, at least 60% idle): Tasks: 296 total, 3 running, 292 sleeping, 0 stopped, 1 zombie %Cpu0 : 14,6 us, 0,7 sy, 0,0 ni, 84,7 id, 0,0 wa, 0,0 hi, 0,0 si, 0,0 st %Cpu1 : 11,2 us, 1,4 sy, 0,0 ni, 87,5 id, 0,0 wa, 0,0 hi, 0,0 si, 0,0 st %Cpu2 : 14,5 us, 1,3 sy, 0,0 ni, 84,2 id, 0,0 wa, 0,0 hi, 0,0 si, 0,0 st %Cpu3 : 4,4 us, 0,7 sy, 0,0 ni, 94,9 id, 0,0 wa, 0,0 hi, 0,0 si, 0,0 st %Cpu4 : 39,6 us, 1,3 sy, 0,0 ni, 59,1 id, 0,0 wa, 0,0 hi, 0,0 si, 0,0 st %Cpu5 : 2,4 us, 0,7 sy, 0,0 ni, 96,6 id, 0,3 wa, 0,0 hi, 0,0 si, 0,0 st %Cpu6 : 2,7 us, 4,3 sy, 0,0 ni, 93,0 id, 0,0 wa, 0,0 hi, 0,0 si, 0,0 st %Cpu7 : 21,0 us, 0,7 sy, 0,0 ni, 78,3 id, 0,0 wa, 0,0 hi, 0,0 si, 0,0 st КiB Mem: 12261384 total, 10578352 used, 1683032 free, 4809968 buffers КiB Swap: 7999484 total, 584072 used, 7415412 free. 1431448 cached Mem PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND21245 user 20 0 17,516g 235792 187228 R 101,3 1,9 1:21.15ffmpeg1512 root -51 0 0 0 0 S 7,0 0,0 17:29.39irq/47-nvidia ----------------- Memory: # free -m total used free shared buffers cach ed Mem: 11974 8619 3354 133 3758 653 -/+ buffers/cache: 4207 7766 Swap: 7811 570 7241 ----------------- Storage I/O: # iostat -x 1 3 Linux 3.13.0-142-generic (user-desktop) 28.02.2018 _x86 _64_ (8 CPU) avg-cpu: %user %nice %system %iowait %steal %idle 10,86 3,03 1,43 4,75 0,00 79,93 Device: rrqm/s wrqm/s r/s w/s rkB/s wkB/s avgrq-sz avgqu-sz await r_await w_await svctm %util sda 0,49 0,00 0,01 0,00 0,29 0,00 45 ,90 0,00 47,45 47,45 0,00 14,55 0,02 sdb 849,21 769,16 47,81 22,18 3859,21 3935,51 222 ,74 9,18 131,22 23,62 363,20 4,05 28,38 avg-cpu: %user %nice %system %iowait %steal %idle 14,65 0,00 1,78 0,00 0,00 83,57 Device: rrqm/s wrqm/s r/s w/s rkB/s wkB/s avgrq-sz avgqu-sz await r_await w_await svctm %util sda 0,00 0,00 0,00 0,00 0,00 0,00 0 ,00 0,00 0,00 0,00 0,00 0,00 0,00 sdb 0,00 0,00 0,00 0,00 0,00 0,00 0 ,00 0,00 0,00 0,00 0,00 0,00 0,00 avg-cpu: %user %nice %system %iowait %steal %idle 21,24 0,00 2,53 0,51 0,00 75,73 Device: rrqm/s wrqm/s r/s w/s rkB/s wkB/s avgrq-sz avgqu-sz await r_await w_await svctm %util sda 0,00 0,00 0,00 0,00 0,00 0,00 0 ,00 0,00 0,00 0,00 0,00 0,00 0,00 sdb 0,00 7,00 0,00 3,00 0,00 88,00 58 ,67 0,04 13,33 0,00 13,33 13,33 4,00 -------------------- FFmper version and configuration options: # /usr/local/ffmpeg-dev/bin/ffmpeg -version ffmpeg version N-90054-g474194a Copyright (c) 2000-2018 the FFmpeg developers built with gcc 4.8 (Ubuntu 4.8.4-2ubuntu1~14.04.4) configuration: --prefix=/usr/local/ffmpeg-dev --enable-gpl --enable- nonfree --enable-libfdk-aac --enable-libx264 --enable-nvenc --enable- libnpp libavutil 56. 7.101 / 56. 7.101 libavcodec 58. 11.101 / 58. 11.101 libavformat 58. 9.100 / 58. 9.100 libavdevice 58. 1.100 / 58. 1.100 libavfilter 7. 12.100 / 7. 12.100 libswscale 5. 0.101 / 5. 0.101 libswresample 3. 0.101 / 3. 0.101 libpostproc 55. 0.100 / 55. 0.100 --------------------- NVIDIA driver and card information: # nvidia-smi Wed Feb 28 17:20:13 2018 +-------------------------------------------------------------------- ---------+ | NVIDIA-SMI 384.111 Driver Version: 384.111 | |-------------------------------+----------------------+--------------- -------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU- Util Compute M. | |===============================+======================+=============== =======| | 0 GeForce GTX 106... Off | 00000000:01:00.0 On | N/A | | 0% 53C P2 32W / 150W | 939MiB / 6071MiB | 11% Default | +-------------------------------+----------------------+--------------- -------++-----------------------------------------------------------------------------+ | Processes: GPU Memory | | GPU PID Type Process name Usage | |====================================================================== =======| | 0 1423 G /usr/bin/X 443MiB | | 0 2555 G compiz 224MiB | | 0 14879 G ...-token=ACEXXXXXXXXX XX2E9DDXXXXXXXE41 107MiB | | 0 21458 C /usr/local/ffmpeg- dev/bin/ffmpeg 159MiB | +-------------------------------------------------------------------- ---------+ I believe I overlooked something, or maybe there are some limitations. So, I kindly ask your suggestions. Many thanks in advance! Garri _______________________________________________ ffmpeg-user mailing list [email protected] http://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email [email protected] with subject "unsubscribe".
<<attachment: marcin_wozniak.vcf>>
_______________________________________________ ffmpeg-user mailing list [email protected] http://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email [email protected] with subject "unsubscribe".
