Question Handbrake 1.3.3 - Benchmark your System - New benchmark criteria

Page 2 - 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
4,191
1,975
136
Back in 2018 ElFenix posted the original Handbrake Benchmark Thread. I thought it was cool so I started updating the thread with a results table. With his permission I'm posting an update to that thread. If you want to participate we're going to tighten up the testing methodology (see below) so the results provide data better for analysis.

I'll update the second post of this thread periodically with results.

The test file:
sdifox said:
et voila. we can all host the file and round robin it :p

LG New York HDR UHD 4K Demo.zip

drive.google.com
drive.google.com


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 average effective clockspeed using HWinfo - https://www.hwinfo.com/
I'm sure there are better ways to do this using the logging feature but here's how I did it. I downloaded the portable version, which runs from the exe file.

Run the exe and select "sensors only." You only need this window open.
Minimize the sensors window so you can have it and Handbrake visible at the same time. Select the "average effective clockspeed" line so you can spot it easily. It will highlight.
Note the icon for "reset min/max/average values and timer" in the bottom right area of the window, it's the clock icon.

Start the handbrake but don't start/reset the timer until handbrake is actually encoding video. Record the average effective clockspeed right before handbrake finishes the encode. It should settle in pretty quickly after the encode starts and not change during the encode.

More info on "effective" vs "instant" clock: https://www.hwinfo.com/forum/threads/effective-clock-vs-instant-discrete-clock.5958/

3. Report "Total CPU Usage" (also from HWinfo)

4. Report your encoding time, CPU Model, and RAM specs (amount, speed, latency)

1665102318265.png
 
Last edited:
  • Like
Reactions: Kaluan and ElFenix

Justinus

Diamond Member
Oct 10, 2005
3,167
1,509
136
Got the Optimus block running - Redid the overclock and test.

5950x
4.725 GHz all core overclock
3.960 GHz average effective clockspeed
3800CL16 1:1

encoded 1806 frames in 118.32s (15.26 fps), 11820.04 kb/s, Avg QP:29.09
 

tamz_msc

Diamond Member
Jan 5, 2017
3,725
3,554
136
i5-1135G7
2447 MHz average effective clockspeed
99.2% average cpu suage
DDR4-2133CL15
encoded 1806 frames in 850.77s (2.12 fps), 11799.19 kb/s, Avg QP:29.12
 
Last edited:

mikk

Diamond Member
May 15, 2012
4,112
2,108
136
CPU: 11th Gen Intel(R) Core(TM) i7-1165G7 @ 2.80GHz
2x16GB DDR4-3200 22-22-22-52
work: average encoding speed for job is 2.353729 fps
encoded 1806 frames in 775.01s (2.33 fps), 11820.04 kb/s, Avg QP:29.09
2712 Mhz effective CPU clock speed

It does not use AVX512 by the way, Tigerlake could perform like 3-4% faster using AVX512.
 

tamz_msc

Diamond Member
Jan 5, 2017
3,725
3,554
136
CPU: 11th Gen Intel(R) Core(TM) i7-1165G7 @ 2.80GHz
2x16GB DDR4-3200 22-22-22-52
work: average encoding speed for job is 2.353729 fps
encoded 1806 frames in 775.01s (2.33 fps), 11820.04 kb/s, Avg QP:29.09
2712 Mhz effective CPU clock speed

It does not use AVX512 by the way, Tigerlake could perform like 3-4% faster using AVX512.
Interesting to note that your average effective clock speed went below the base frequency while mine didn't.
 

mikk

Diamond Member
May 15, 2012
4,112
2,108
136
Interesting to note that your average effective clock speed went below the base frequency while mine didn't.

The effective clock speed is a bit under the base because of the first and last seconds when it isn't fully loaded and the CPU utilization overall is like 98-99%. If you deactivate your CPU Turbo your effective CPU speed should be a bit lower than the base. I did disable the CPU Turbo because it's easier for a stable performance testing.
 

tamz_msc

Diamond Member
Jan 5, 2017
3,725
3,554
136
The effective clock speed is a bit under the base because of the first and last seconds when it isn't fully loaded and the CPU utilization overall is like 98-99%. If you deactivate your CPU Turbo your effective CPU speed should be a bit lower than the base. I did disable the CPU Turbo because it's easier for a stable performance testing.
How do you disable turbo boost? I don't think my BIOS has an option to do that.
 

mikk

Diamond Member
May 15, 2012
4,112
2,108
136
CPU: 11th Gen Intel(R) Core(TM) i7-1165G7 @ 2.80GHz
2x16GB DDR4-3200 22-22-22-52
work: average encoding speed for job is 2.353729 fps
encoded 1806 frames in 775.01s (2.33 fps), 11820.04 kb/s, Avg QP:29.09
2712 Mhz effective CPU clock speed

It does not use AVX512 by the way, Tigerlake could perform like 3-4% faster using AVX512.


Actually the AVX512 speedup is a bit higher in this test.

AVX512 (asm=avx512)
CPU: 11th Gen Intel(R) Core(TM) i7-1165G7 @ 2.80GHz
2x16GB DDR4-3200 22-22-22-52
work: average encoding speed for job is 2.516485 fps
encoded 1806 frames in 722.67s (2.50 fps), 11820.04 kb/s, Avg QP:29.09
2700 Mhz effective CPU clock speed
Total CPU Usage ~96.2% average
 
Last edited:

Hulk

Diamond Member
Oct 9, 1999
4,191
1,975
136
tamz_msc and mikk

What laptop models do you guys have? I think it might be nice to have that info in the chart.

mikk,

I'm not quite following the avx thing. Is there a way to enable this in Handbrake? Could you bring me up to speed on this?
 

mikk

Diamond Member
May 15, 2012
4,112
2,108
136
tamz_msc and mikk

What laptop models do you guys have? I think it might be nice to have that info in the chart.

mikk,

I'm not quite following the avx thing. Is there a way to enable this in Handbrake? Could you bring me up to speed on this?


Yes it can be enabled in the video tab with the parameter asm=avx512

It's disabled by default, at least on Tigerlake. I have a Dell Inspiron 15 7506.
 

tamz_msc

Diamond Member
Jan 5, 2017
3,725
3,554
136
tamz_msc and mikk

What laptop models do you guys have? I think it might be nice to have that info in the chart.

mikk,

I'm not quite following the avx thing. Is there a way to enable this in Handbrake? Could you bring me up to speed on this?
I'll do a run later today with AVX512 enabled. My laptop model is a HP 14s-dr2006tu.
 

tamz_msc

Diamond Member
Jan 5, 2017
3,725
3,554
136
Looks like AVX512 gives worse performance due to downclocking:

encoded 1806 frames in 899.81s (2.01 fps), 11820.04 kb/s, Avg QP:29.09
Average effective clock speed: 1970 MHz
Total CPU usage: 89% average

But on the flipside IPC or in this case PPC(performance per core) per gigahertz is 0.255, which I believe is the highest so far.

Edit: @Hulk could you add my and @mikk's AVX512 scores as a separate bar in the chart?
 
Last edited:

dr1337

Senior member
May 25, 2020
311
516
106
Ive been tweaking my work rig today so I figured I'd give this a go for funsies

5600x all stock with the stock cooler
16gb cl16 running at 3800mhz
Avg. Effective clock 3798 mhz
Avg. total CPU usage 96.8%
Results: encoded 1806 frames in 330.50s (5.46 fps), 11820.04 kb/s, Avg QP:29.09
 

Hulk

Diamond Member
Oct 9, 1999
4,191
1,975
136
Looks like AVX512 gives worse performance due to downclocking:

encoded 1806 frames in 899.81s (2.01 fps), 11820.04 kb/s, Avg QP:29.09
Average effective clock speed: 1970 MHz
Total CPU usage: 89% average

But on the flipside IPC or in this case PPC(performance per core) per gigahertz is 0.255, which I believe is the highest so far.

Edit: @Hulk could you add my and @mikk's AVX512 scores as a separate bar in the chart?

Done.
Your efficiency score seems really high compared to the other Tiger Lake and my Kaby Lake R w/wo AVX512. Can you double check that?

And yes, AVC512 loads the heck out of the cores and for laptop systems, which are already thermally constrained the situation gets worse.
I took my Surface Laptop 2 into my 40 degree F garage and the clock went up a few hundred MHz. Then I enabled AVX but couldn't hold the same clock so my score ended up about the same.
My Kaby Lake R is 5% more efficient with this test using AVX512.

One thing I'm wondering is the value in purchasing products higher in the stack for laptops, which as I wrote above are thermally constrained.

Let's consider the following CPUs.
downupmax
1135G79002.404.20
1165G71.22.804.70

In a laptop that can let's say dissipate 15W how much faster under full load with all cores loaded will the 1165 operate vs the 1165? Or will they both be constrained at the same point?
Now I realize that the higher binned 1165 should have a voltage/frequency curve that allows for higher frequency at lower voltage and hopefully power but who knows if you are really getting what you are paying for?

I would assume the single core clocks are more attainable since less power overall needs to be dissipated. But again no one seems to test this with two identical laptops, one with fastest CPU and one with slowest.
 

tamz_msc

Diamond Member
Jan 5, 2017
3,725
3,554
136
@Hulk, the efficiency is high because the clock speed is so low, though you're right I should double-check and will probably do so if time permits.

The main reason for downclocking is the fact that this is an all-core load, which is not what you'd want while running AVX-512, and the fact that tau on my laptop is set pretty aggressively. Too bad I can't disable Turbo Boost like @mikk can on his Inspiron.
 

Hulk

Diamond Member
Oct 9, 1999
4,191
1,975
136
@Hulk,
The main reason for downclocking is the fact that this is an all-core load, which is not what you'd want while running AVX-512, and the fact that tau on my laptop is set pretty aggressively. Too bad I can't disable Turbo Boost like @mikk can on his Inspiron.

I'm sorry to be a drag but I'm not totally following here. Can you break it down for me a bit more please?
 

mikk

Diamond Member
May 15, 2012
4,112
2,108
136
And yes, AVC512 loads the heck out of the cores and for laptop systems, which are already thermally constrained the situation gets worse.
I took my Surface Laptop 2 into my 40 degree F garage and the clock went up a few hundred MHz. Then I enabled AVX but couldn't hold the same clock so my score ended up about the same.
My Kaby Lake R is 5% more efficient with this test using AVX512.


Kabylake shouldn't be there in the AVX512 list because it does not support AVX512, it's just for Icelake-U, Tigerlake-U or Skylake-X. You can check this in the log:

x265 [info]: using cpu capabilities: MMX2 SSE2Fast LZCNT SSSE3 SSE4.2 AVX FMA3 BMI2 AVX2

This is what you should get on a Skylake based device. And this is Tigerlake-U:

x265 [info]: using cpu capabilities: MMX2 SSE2Fast LZCNT SSSE3 SSE4.2 AVX FMA3 BMI2 AVX2 AVX512
 

Hulk

Diamond Member
Oct 9, 1999
4,191
1,975
136
Kabylake shouldn't be there in the AVX512 list because it does not support AVX512, it's just for Icelake-U, Tigerlake-U or Skylake-X. You can check this in the log:

x265 [info]: using cpu capabilities: MMX2 SSE2Fast LZCNT SSSE3 SSE4.2 AVX FMA3 BMI2 AVX2

This is what you should get on a Skylake based device. And this is Tigerlake-U:

x265 [info]: using cpu capabilities: MMX2 SSE2Fast LZCNT SSSE3 SSE4.2 AVX FMA3 BMI2 AVX2 AVX512

I know AVX2 is basically half the speed of AVX512 so that's what's going on here right?

x265 [info]: using cpu capabilities: MMX2 SSE2Fast LZCNT SSSE3 SSE4.2 AVX FMA3 BMI2 AVX2
 
Last edited:

tamz_msc

Diamond Member
Jan 5, 2017
3,725
3,554
136
I'm sorry to be a drag but I'm not totally following here. Can you break it down for me a bit more please?
Tau refers to the amount of time the CPU spends in PL2 before dropping to PL1. It is a part of Intel's TDP spec and is set by OEMs depending on the thermal capabilities of the laptop/desktop design. In DIY space motherboards, especially high end ones, typically ignore it, allowing you to run at PL2 all the time as required.
 
  • Like
Reactions: Hulk

Noid

Platinum Member
Sep 20, 2000
2,373
183
106
85.9 percent
encoded 1806 frames in 170.10s (10.62 fps), 11820.04 kb/s, Avg QP:29.09

5900X
2x16 32GB 3800 Cas18

This is BOOST mode.

I think an All Core test should be next.
 

Attachments

  • Capture.JPG
    Capture.JPG
    173.1 KB · Views: 8
Last edited:

Hulk

Diamond Member
Oct 9, 1999
4,191
1,975
136
85.9 percent
encoded 1806 frames in 170.10s (10.62 fps), 11820.04 kb/s, Avg QP:29.09

5900X
2x16 32GB 3800 Cas18

This is BOOST mode.

As you can see in my screenshot avg core speed is 4.2 GHz.

I think an All Core test should be next.

Thanks for submitting your score. We're reporting average effective clock not average clock. Reset just after Handbrake starts encoding and then report the average effective clock value right before it finishes encoding. Would you mind doing this?
 

Noid

Platinum Member
Sep 20, 2000
2,373
183
106
encoded 1806 frames in 170.24s (10.61 fps), 11820.04 kb/s, Avg QP:29.09

Average Effective Clock screenshot at 93% done.

Which " average " column number you want ... is questionable....
Averages ... of averages ... are not accurate.

This test is not pushing the limits of this chip.

I'm expecting an ALL core OC of 4.65 GHz
( not testing yet )
Will submit when All Core found.

( this is a BOOST score )
 

Attachments

  • Capture.JPG
    Capture.JPG
    229.9 KB · Views: 11
Last edited: