Question Handbrake 1.3.3 - Benchmark your System - COMPLETE Overhaul of the test

Page 15 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Hulk

Diamond Member
Oct 9, 1999
5,048
3,529
136
A little background...
Handbrake is a ubiquitous encoding application and happens to be one that makes good use of multicore/thread CPU's when encoding x265. x265 is a widely used and efficient compression scheme that requires significant compute to encode. While hardware encoders are faster, at the same bitrates, CPU (software) encode produces better video quality. Of course this assumes the use of lower bitrates as quality for both hardware and software encodes will be indistinguishable at higher bitrates. But the point of the video encode is to get good quality at low bitrates so we are therefore testing software encode.

fps/GHz/core is a representation of how efficient a given CPU core is at encoding the test file using the x265 format. The number is arrived at by multiplying the number of physical cores by the average frequency they are running at and then dividing by the fps from the Handbrake test. It tells us for a given core how many fps can this core encode the test if it was running at 1GHz. We could consider this an "IPC" of sorts for this test but strictly speaking this would be closer to the word "throughput." And as you know many around here are indeed strict with terminology so I will avoid the word IPC at it denotes Instructions Per Cycle and that is not actually what we are measuring.

Some people will go "all out" and try and run their system as close to the limit as possible and others (like me) just run at stock. All of the data is valuable and informative as long as it is collected from each person in the same manner and there for comparable.

I went through all of the results and created a new table. In respecting everyone's time who participated in the old data I am keeping that data on the 2nd page of this post.

Here's the test file: https://4kmedia.org/lg-new-york-hdr-uhd-4k-demo/


1. Use the following version of Handbrake with the built-in h.265 mkv 2160p60 preset
HandBrake-1.3.3-x86_64-Win_GUI.exe
Don't forget to turn on logging in Handbrake so you can retrieve your time. Tools>Preferences>Advanced>Logging
Once this current version is replaced you'll be able to access this version from the following link.
HandBrake: Nightly Builds
Nightly builds of HandBrake
handbrake.fr

2. Report your encoding time, average CPU frequency, and Package Power. If you have a hybrid CPU you can turn off the E's in the BIOS. For E testing turn off all P's except one in the BIOS, clock it down to 800MHz, and then shut it down with Process Lasso. Or just report your score with 1 P at 800MHz and let me know you did that so I can subtract out that P core's (minor) contribution to the encode.

Here's how to report your average clock and package power so we are all doing it the same way.
Handbrake does some housekeeping right after you start encode and when the progress bar gets to 100%.
This could result in lower than actual average clock.
After you start the encode, wait a few seconds until you see the green Handbrake bar appear, then reset the HWinfo counter.
At the end don't wait to grab the screen shot at 100%, just do it sometime after about 95%.

3. CPU Model, and RAM specs
 

Attachments

  • Handbrake.chart.jpg
    581.4 KB · Views: 33
  • Handbrake.new.jpg
    1.2 MB · Views: 23
Last edited:

ArrogantHair

Member
Feb 6, 2025
26
14
36
My 285K P54E46
900 frames/19.07fps = 47.19s
My Ryzen 5 3600
900 frames/4.11fps = 218.98s

The 3600 is a media server with services running 16GB DDR4 - not optimal for test.
I have a Windows server with a 5700x and 64GB DDR4, but it has issues with Handbrake.

Igor's 245KF
6P+8E 900 frames/12.05fps = 74.69s
6P+4E 900 frames/9.46fps = 95.14s
6P+2E 900 frames/8.25fps = 109.9s
6P+1E 900 frames/7.55fps = 119.21s
6P+0E 900 frames/6.84fps = 131.58s

 
Reactions: igor_kavinski

ArrogantHair

Member
Feb 6, 2025
26
14
36
so I did some runs with my 9950X. These are the results according to your encoding calculation method:

Stock 200W PPT / DDR5 6000: 47,39 s
View attachment 117062

PBO Tuned 180W PPT / DDR5 6400: 43,85 s
View attachment 117057

PBO Tuned 200W PPT / DDR5 6400: 43,10 s
View attachment 117058

PBO Tuned 230W PPT / DDR5 6400: 42,86 s
View attachment 117059

PBO Tuned 250W PPT / DDR5 6400: 42,35 s
View attachment 117060

Summary (including your 285K results):

View attachment 117064
That looks good. My original encoding was to MKV instead of MP4 which is slightly slower. I re-ran for P54E46 for MP4 to get 47.19s.
My tuning and voltages are still off for my overclocks - redoing.
 
Reactions: lightmanek

MarkPost

Senior member
Mar 1, 2017
374
788
136
That looks good. My original encoding was to MKV instead of MP4 which is slightly slower. I re-ran for P54E46 for MP4 to get 47.19s.
My tuning and voltages are still off for my overclocks - redoing.
well 47.39 mkv vs 47.19 mp4, is 0.42% difference, thats basically the same time.

I've noticed CPU utilization is really poor: below 70% average during encoding. It seems that this preset isn't the best optimized or video source isn't the right one in this regard.
 

Hulk

Diamond Member
Oct 9, 1999
5,048
3,529
136
well 47.39 mkv vs 47.19 mp4, is 0.42% difference, thats basically the same time.

I've noticed CPU utilization is really poor: below 70% average during encoding. It seems that this preset isn't the best optimized or video source isn't the right one in this regard.
mkv and mp4 are just containers for the data, x265 can be in either one.
 

MS_AT

Senior member
Jul 15, 2024
533
1,124
96
Wouldn't know how to make the necessary changes to build with Clang.
Actually if you are happy with cli, then you can try to use the one provided by msys2 https://www.msys2.org/ they provide clang compiled version of handbrake https://packages.msys2.org/packages/mingw-w64-clang-x86_64-handbrake but without digging deeper it's hard to say which options they used for compiling. Before using, if you are unfamiliar with msys read up about them on their page (first link).
 
Reactions: igor_kavinski

Hulk

Diamond Member
Oct 9, 1999
5,048
3,529
136
Looking into Handbrake's behavior as cores from the 9950X are added. If you start the encode and pause like 2 seconds later, right before when the actual encode starts, you can select the cores you want to work on the encode from Task Manager>Details/Handbrake (select)>right click Set Affinity.

There is a glitch in performance at 14 cores/28 threads. Tested that data point 3 times with the same result. I have no explanation other than my speculation that something in the code doesn't fit the available hardware resources as well as the other data points.

Something interesting to note, max power of about 195W is hit at 12 cores but encode framerate keeps increasong as you move to 14 and 16 cores even as average clock speed decreases. This is of course what we would expect considering the nonlinear nature of the end of the v/f curve. The power is more efficiently used clocking more cores higher rather than less cores higher.

Additional silicon area is useful for MT, but also commensurately expensive.

As Intel has told us the E's are area efficiency for MT. Yes, they make better use of power for MT as well, but the reason they are they is to preserve MT performance while conserving die space.


 
Last edited:
Reactions: lightmanek

Nothingness

Diamond Member
Jul 3, 2013
3,249
2,304
136

jamesdsimone

Senior member
Dec 21, 2015
884
245
116
well 47.39 mkv vs 47.19 mp4, is 0.42% difference, thats basically the same time.

I've noticed CPU utilization is really poor: below 70% average during encoding. It seems that this preset isn't the best optimized or video source isn't the right one in this regard.
I noticed the same thing. It usually hovers between 60-70%. If you run 2 encodes at the same time the CPU utilization goes all the way to 100%.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |