Benchmark your computer with Handbrake 1.01 and x265

[Zucker2k]

encoded 1497 frames in 47.86s (31.28 fps), 4036.02 kb/s, Avg QP:26.24
HandBrake 1.0.7 (2017040900) - 64bit
OS: Microsoft Windows NT 10.0 - 64bit
CPU: Intel(R) Core(TM) i9-7980XE CPU @ 4.60GHz
Ram: 32GB, DDR4 - 3200 (12-12-18-28 1t)
GPU Information:
NVIDIA GTX 670
Screen: 1920x1080

Hotdamn!

 

[Paratus]

encoded 1497 frames in 400.11s (3.74 fps), 4036.02 kb/s, Avg QP:26.24
Handbrake 1.01
OS: Windows 10 Pro
CPU: I7 920 @3.5GHz
RAM: 12GB, DDR3 1333 (8-9-9-24-74-2T)
GPU: MSI Lightning 7970 GHZ Ed @1070/5600
Screen: 2560x1440 @ 60P

Tremble at my mighty i7 920! Only 8.4 times slower than Totalnet

 

[kjboughton]

Good run here:

encoded 1497 frames in 60.24s (24.85 fps), 4036.02 kb/s, Avg QP:26.24
HandBrake 1.0.1 (2016122900) - 64bit
OS: Windows 10 Enterprise (Version 1709, Build 16299.19)
CPU: Dual Intel Xeon E5-2696 v3 (Overclocked)
Ram: 64GB, DDR4-2133R
GPU: NVIDIA GeForce GTX 1080 Ti
Screen: 3440x1440

 

[psolord]

Here is my Core i5-8600k@5Ghz result. RAM 3200-14-14-14-35

encoded 1497 frames in 121.25s (12.35 fps), 4036.02 kb/s, Avg QP:26.24

Only I'm using handbrake 1.07.

 

[palladium]

encoded 1497 frames in 75.31s (19.88 fps), 4036.02 kb/s, Avg QP:26.24

i9 7920x @ 4.2 (-400 AVX offset); 32GB DDR4 2400. Solid 100% CPU usage with occasional brief blips to 90%

 

[Justinus]

encoded 1497 frames in 47.86s (31.28 fps), 4036.02 kb/s, Avg QP:26.24
HandBrake 1.0.7 (2017040900) - 64bit
OS: Microsoft Windows NT 10.0 - 64bit
CPU: Intel(R) Core(TM) i9-7980XE CPU @ 4.60GHz
Ram: 32GB, DDR4 - 3200 (12-12-12-28 1t)
GPU Information:
NVIDIA GTX 670
Screen: 1920x1080

I would be extremely interested to know how disabling hyperthreading to present only 18 physical cores to handbrake affects your score. Any chance of a non-HT run?

Also, got a replacement 5960x. 4.7 core, 4.6 cache. DDR4 3200 14-15-15-30 1T

encoded 1497 frames in 93.77s (15.97 fps), 4035.92 kb/s, Avg QP:26.24

I had to run 1.07 because 1.01 crashes every time I try to use it on 1709, even after a reinstall.

 

[Totalnet]

I would be extremely interested to know how disabling hyperthreading to present only 18 physical cores to handbrake affects your score. Any chance of a non-HT run?

Also, got a replacement 5960x. 4.7 core, 4.6 cache. DDR4 3200 14-15-15-30 1T

encoded 1497 frames in 93.77s (15.97 fps), 4035.92 kb/s, Avg QP:26.24

I had to run 1.07 because 1.01 crashes every time I try to use it on 1709, even after a reinstall.

I Will try to do a non HT run this week..

Regards, Dolf

 

[Zucker2k]

I Will try to do a non HT run this week..

Regards, Dolf

Yay!!!

 

[psolord]

Here is my Core i5-8600k@5Ghz result. RAM 3200-14-14-14-35

encoded 1497 frames in 121.25s (12.35 fps), 4036.02 kb/s, Avg QP:26.24

Only I'm using handbrake 1.07.

I did a run on my old core i7 860@4Ghz out of curiosity.

encoded 1497 frames in 336.25s (4.45 fps), 4036.02 kb/s, Avg QP:26.24

The 8600k@5Ghz is 2.78 times faster. 2.78 X i7 860 that is, because I had a big argument in another forum if one times faster is actually a correct expression and what it really means.

The silly thing actually hit 99C. wtf? I have never seen the 860 getting so hot. Ok I will dismantle it next week anyway, but still.

 

[Totalnet]

encoded 1497 frames in 47.96s (31.21 fps), 4036.34 kb/s, Avg QP:26.24
HandBrake 1.0.7 (2017040900) - 64bit
OS: Microsoft Windows NT 10.0 - 64bit
CPU: Intel(R) Core(TM) i9-7980XE CPU @ 4.60GHz - HT=OFF
Ram: 32GB, DDR4 - 3200 (12-12-12-28 1t)
GPU Information:
NVIDIA GTX 670
Screen: 1920x1080
HT has no effect at all, same score..

 

[Totalnet]

encoded 1497 frames in 46.49s (32.20 fps), 4036.02 kb/s, Avg QP:26.24
HandBrake 1.0.7 (2017040900) - 64bit
OS: Microsoft Windows NT 10.0 - 64bit
CPU: Intel(R) Core(TM) i9-7980XE CPU @ 4.80GHz - HT=OFF
Ram: 32GB, DDR4 - 3200 (12-12-12-28 1t)
GPU Information:
NVIDIA GTX 670
Screen: 1920x1080

Without HT i can do a 4.8Ghz run

 

[Justinus]

encoded 1497 frames in 47.96s (31.21 fps), 4036.34 kb/s, Avg QP:26.24
HandBrake 1.0.7 (2017040900) - 64bit
OS: Microsoft Windows NT 10.0 - 64bit
CPU: Intel(R) Core(TM) i9-7980XE CPU @ 4.60GHz - HT=OFF
Ram: 32GB, DDR4 - 3200 (12-12-12-28 1t)
GPU Information:
NVIDIA GTX 670
Screen: 1920x1080
HT has no effect at all, same score..

encoded 1497 frames in 46.49s (32.20 fps), 4036.02 kb/s, Avg QP:26.24
HandBrake 1.0.7 (2017040900) - 64bit
OS: Microsoft Windows NT 10.0 - 64bit
CPU: Intel(R) Core(TM) i9-7980XE CPU @ 4.80GHz - HT=OFF
Ram: 32GB, DDR4 - 3200 (12-12-12-28 1t)
GPU Information:
NVIDIA GTX 670
Screen: 1920x1080

Without HT i can do a 4.8Ghz run

I had a feeling 18 cores was around the point that the failure to scale to higher thread counts would have HT potentially end up breaking even or possibly resulting in a performance loss. Great results and very informative.

 

[ElFenix]

you guys want to run 2 instances and average them?


edit: should probably just try 4k recoding. new tab on the spreadsheet time!

 

[StinkyPinky]

It seems your cpu is throttling or something is wrong. Its only ~9% faster than Crono's 8700 at 4.3

EDIT: ummm or is it AVX offset? about -300 or so?

Sorry I missed this. The answer is yes it has an offset of exactly 300 actually.

 

[Zucker2k]

encoded 1497 frames in 46.49s (32.20 fps), 4036.02 kb/s, Avg QP:26.24
HandBrake 1.0.7 (2017040900) - 64bit
OS: Microsoft Windows NT 10.0 - 64bit
CPU: Intel(R) Core(TM) i9-7980XE CPU @ 4.80GHz - HT=OFF
Ram: 32GB, DDR4 - 3200 (12-12-12-28 1t)
GPU Information:
NVIDIA GTX 670
Screen: 1920x1080

Without HT i can do a 4.8Ghz run

Nice. Would you consider running the cinebench 11.5 challenge? Would be nice to see how it compares to threadripper there as well. You have a monster on your hands.

 

[Totalnet]

Nice. Would you consider running the cinebench 11.5 challenge? Would be nice to see how it compares to threadripper there as well. You have a monster on your hands.

Posted my result..
It's a Threadripper-ripper ...

 

[palladium]

encoded 1497 frames in 47.86s (31.28 fps), 4036.02 kb/s, Avg QP:26.24
HandBrake 1.0.7 (2017040900) - 64bit
OS: Microsoft Windows NT 10.0 - 64bit
CPU: Intel(R) Core(TM) i9-7980XE CPU @ 4.60GHz
Ram: 32GB, DDR4 - 3200 (12-12-12-28 1t)
GPU Information:
NVIDIA GTX 670
Screen: 1920x1080

How much Vcore did you need to get 4.6?

 

[Totalnet]

How much Vcore did you need to get 4.6?

Ik think it was 1.30v

 

[Evidences]

Stumbled across this thread trying to find any info on encoding times for h265 since I'm planning on doing a decent amount of it in the future So since I found this thread somewhat useful I figured I'd throw a few test out for you guys and anyone else who many stumble across this thread in the future.

I didn't think to grab screenshots or anything for these until I got a couple tests in but I can assure you I'm not making any of these numbers up and all of them are 100% legit. Also I'm willing to give any more info anyone wants upon request.

I'll start with the quad cores and move on from here.

Cpu I5-2500k @ 4.0GHz ddr3-1600 Windows 10 pro
encoded 1497 frames in 362.82s (4.13 fps), 4036.02 kb/s, Avg QP:26.24

Cpu AMD Phenom II X4 940 @ 3.0GHz ddr2-800 Windows 10 pro
encoded 1497 frames in 1409.23s (1.06 fps), 4036.02 kb/s, Avg QP:26.24

These are dual core cpus though the i3 does have hyper-threading so make of it what you will.

Cpu i3-7100U CPU @ 2.40GHz ddr4-2133 Windows 10 home
encoded 1497 frames in 621.16s (2.41 fps), 4036.02 kb/s, Avg QP:26.24

Cpu AMD Athlon 64 X2 4400+ @ 2.3GHz ddr2-800 Windows 10 pro
encoded 1497 frames in 5151.28s (0.29 fps), 3978.84 kb/s, Avg QP:26.14

Decided this was too slow so I overclocked this chip till it started failing in prime95.

Cpu AMD Athlon 64 X2 4400+ @ 3.1GHz ddr2-800 Windows 10 pro
encoded 1497 frames in 3714.00s (0.40 fps), 3978.81 kb/s, Avg QP:26.14

The rest of these are all socket 1366 Xeons, I tested all of these in both single and dual socket configurations when possible just to be thorough.

Cpu single X5687 @ 3.60GHz ddr3-1066 Windows 10 pro
encoded 1497 frames in 360.45s (4.15 fps), 4036.02 kb/s, Avg QP:26.24

Cpu single L5640 @ 2.27GHz ddr3-1066 Windows 10 pro
encoded 1497 frames in 401.25s (3.73 fps), 4036.02 kb/s, Avg QP:26.24

Cpu dual L5640 @ 2.27GHz ddr3-1066 Windows 10 pro
encoded 1497 frames in 232.62s (6.44 fps), 4036.02 kb/s, Avg QP:26.24

Cpu Single X5670 @ 2.93GHz ddr3-1066 Windows 10 pro
encoded 1497 frames in 309.44s (4.84 fps), 4036.02 kb/s, Avg QP:26.24

Cpu dual X5670 @ 2.93GHz ddr3-1066 Windows 10 pro
encoded 1497 frames in 182.66s (8.20 fps), 4036.02 kb/s, Avg QP:26.24

I actually have a couple more CPUs I wanted to test but handbrake doesn't seem to want to run on the AMD K6-3 I have and I unfortunately don't have a working motherboard or ram for the socket 939 CPUs I have lying around.

 

[PeterScott]

encoded 1497 frames in 720.48s (2.08 fps), 4036.02 kb/s, Avg QP:26.24

HandBrake 1.0.7 (2017040900) - 64bit
OS: Microsoft Windows NT 6.1.7601 Service Pack 1 - 64bit
CPU: Intel(R) Core(TM)2 Quad CPU Q9400 @ 3.2GHz (OC)
Ram: 6142 MB,

I did it again to check for a Meltdown Patch impact, and it was faster this time. Though I think I might have ran it on HDD last time, SSD this time:
encoded 1497 frames in 687.01s (2.18 fps), 4036.02 kb/s, Avg QP:26.24

 

[BrandonT]

I did it again to check for a Meltdown Patch impact, and it was faster this time. Though I think I might have ran it on HDD last time, SSD this time:
encoded 1497 frames in 687.01s (2.18 fps), 4036.02 kb/s, Avg QP:26.24

Not apples to apples, but thanks for posting anyway. I hope more people do this comparison so I can get a better idea if I'm being impacted (haven't had time to do the test myself)

 

[Paratus]

On Handbrake 1.1
CPU 1900X stock ~3.9G
MOBO MSI X399 Gamming Pro Carbon
RAM 3200 14-14-14

encoded 1497 frames in 126.44s (11.84 fps), 4626.29 kb/s, Avg QP:26.00

So 3X faster than my OC’d i7920

 

[Bushwicktrini]

HandBrake 1.1.0 (2018040700) - 64bit
OS: Microsoft Windows NT 10.0.16299.0 - 64bit
CPU: Intel(R) Core(TM) i5-8600K CPU @ 5.20GHz
ASUS ROG MAXIMUS X HERO (WI-FI AC)
Ram: 32GB DDR4
GPU Information: NVIDIA GeForce GTX 1070 - 23.21.13.9135

encoded 1497 frames in 127.65s (11.73 fps), 4637.89 kb/s, Avg QP:26.00

 

[richaron]

Latest BIOS let me bump RAM up to 3000 easily so I thought I bench with handbrake... With no difference vs 2800. But I thought I'd put this out there as another data point.

OS: Ubuntu 16.04.4 LTS - Kernel 4.13.0-38-generic
CPU: Ryzen 7 1700 Stock (3.0GHz all core)
RAM: 64GB ECC @ 3000 (CL17)
Mobo: ASUS STRIX X370-F (V. 3807)
HandBrake 1.1.0 (nightly 20180411201741-7cf1ae1-master (x86_64))

encoded 1497 frames in 148.91s (10.05 fps), 4637.28 kb/s, Avg QP:26.01

 

[ElFenix]

1.1 came out, was waiting for it. seems like it may be time to end this thread and start a new one with 1.1 and 4k transcoding since new CPUs loaf at this 4k to 1080p stuff.