AMD Ryzen 5900X Results

Post your testing results with HandBrake.
Post Reply
User avatar
s55
HandBrake Team
Posts: 10347
Joined: Sun Dec 24, 2006 1:05 pm

AMD Ryzen 5900X Results

Post by s55 »

Hardware:
  • CPU: AMD Ryzen 9 5900X 12 Core / 24 Threads
  • Overclock: None
  • ]Memory: 32GB DDR4 3200Mhz
  • Operating System: Windows 10 20H2
Source: bbb_sunflower_1080p_60fps_normal.mp4

Fast 1080p30: 321.15 fps (100% utilisation)
H.265 MKV 1080p30: 86.756 fps (70 ~ 70% utilisation)

Extra:
Fast 1080p30 (No Decomb/Interlace Detect): 433.12 fps (80 ~ 100% utilisation) (Decomb/ID Penalty: 25.8% )


Source: Tears Of Steel 2160p24 FLAC AC3.mkv

Fast 1080p30: 66.52 fps (75% ~ 85% utilisation)
Fast 1080p30 (No Decomb/Interlace Detect): 80.13 fps (~50% utilisation)
H.264 MKV 2160p60: 42.27 fps (~ 100% utilisation)
H.265 MKV 2160p60: 11.11 fps (85 ~ 95% utilisation)
Attachments
Tears_mkv2160_h265.mkv.txt
(10.65 KiB) Downloaded 237 times
Tears_mkv2160_h264.mkv.txt
(10.32 KiB) Downloaded 224 times
BBB_mkv1080p30_h265.mkv.txt
(10.42 KiB) Downloaded 244 times
BBB_fast1080p30.mp4.txt
(10.14 KiB) Downloaded 215 times
rollin_eng
Veteran User
Posts: 4836
Joined: Wed May 04, 2011 11:06 pm

Re: AMD Ryzen 5900X Results

Post by rollin_eng »

Interesting, any thoughts on why utilisation drops to 50% on the Fast encode with 4K source?
User avatar
s55
HandBrake Team
Posts: 10347
Joined: Sun Dec 24, 2006 1:05 pm

Re: AMD Ryzen 5900X Results

Post by s55 »

I'd wager it's limited on decode or filter but I'll need to do some testing to confirm.
rollin_eng
Veteran User
Posts: 4836
Joined: Wed May 04, 2011 11:06 pm

Re: AMD Ryzen 5900X Results

Post by rollin_eng »

Ok makes sense, but do you think reading data from disk? Or ffmpeg decode?
User avatar
s55
HandBrake Team
Posts: 10347
Joined: Sun Dec 24, 2006 1:05 pm

Re: AMD Ryzen 5900X Results

Post by s55 »

It's not IO. It's either ffmpeg decode or possibly filters.

Edit: Not the Decomb/Interlace Detect it seems.
Edit: Corrections to numbers above following re-test. Also seeing quite a bit of run-to-run variance. I may need to do a multi-pass run for each to get averages :(
rollin_eng
Veteran User
Posts: 4836
Joined: Wed May 04, 2011 11:06 pm

Re: AMD Ryzen 5900X Results

Post by rollin_eng »

Is there a reason you can think of that the decode cannot utilise all the cores/cpu?
rollin_eng
Veteran User
Posts: 4836
Joined: Wed May 04, 2011 11:06 pm

Re: AMD Ryzen 5900X Results

Post by rollin_eng »

And why can’t I edit posts anymore???

Is there a technical reason you can think of that the decode cannot utilise all the cores/cpu?
User avatar
s55
HandBrake Team
Posts: 10347
Joined: Sun Dec 24, 2006 1:05 pm

Re: AMD Ryzen 5900X Results

Post by s55 »

Post editing is time limited (prevents spam bots). I've turned it down to be a little less aggressive on lockout.

Short answer. Don't know yet. I'll need to do some digging to see where the issue actually lies.
Post Reply