Benchmark your computer with Handbrake 1.01 and x265

[crashtech]

@VirtualLarry , I wonder how much single channel is impacting your score, this benchmark does involve a lot of main memory transactions.

 

[superstition]

Here is another result for you:

encoded 1497 frames in 809.97s (1.85 fps), 4036.54 kb/s, Avg QP:26.24

“Sandy Bridge” 2.7 GHz i7 2620M dual core (Mac Mini Mid 2011)
256K L2 per core, 4 MB shared L3
8 GB of 1333 MHz DDR3 (2 X 4GB), 750 GB sata mobile HD
macOS 10.12.2

update:

Edgemeal, thanks for changing the construction core PD CPUs to modules. There is an APU in there (rancher lee's A10-7850K) that should be changed from 4C/4T to 2M/4T.

 

[Hulk]

@VirtualLarry , I wonder how much single channel is impacting your score, this benchmark does involve a lot of main memory transactions.

Not at all. Like nearly all video encoding it just wants to eat compute cycles. Newer cores with better IPC do better and clockspeed is king.

In fact I did a little number crunching. I did a calculation I'll call "cycles/fps." That is how many cycles it takes to produce 1fps encoding on this benchmark. I multiplied the number of cores of the CPU times the clockspeed and divided by the encoded fps result from Handbrake. Next I averaged the results of same cores.
Here's what I got:

Skylake - 2084 cycles to produce 1fps encoding speed
Broadwell - 2250 cycles to produce 1fps encoding speed (this is from my t450s laptop so I suspect a desktop system would do a bit better)
Haswell - 2274 cycles to produce 1fps encoding speed
Ivy Bridge - 2935 cycles to produce 1fps encoding speed
Lynnfield - 4318 cycles to produce 1fps encoding speed (result from an i5-750 so no HT on this one)

Can someone bench a 2600k?
Anyone have one of those rare as hens teeth 5775C Broadwell cores to test?

 

[ElFenix]

I just noticed the other OS X result (with the other Macbook Pro) disappeared, too.

looking at the results, i'm seeing both the 750 and the 4850hq in the list, which i haven't edited since last friday. so the results are there, i don't know why it's not showing up for you. cache?

updating list through this post.

 

[VirtualLarry]

My same G4600 @ stock 3.6Ghz with HT, no AVX/AVX2, dual-channel DDR4-2400 @ 2400

encoded 1497 frames in 528.50s (2.83 fps), 4036.02 kb/s, Avg QP:26.24

 

[Conroe]

i7 7700k at 5ghz DDR4-3000 Windows 10. It throttles down to 4.8 sometime. I guess it's the AVX code? Too many things are still set at auto. It did hit 88c so lowering the clock for AVX is a good thing.

encoded 1497 frames in 158.29s (9.46 fps), 4036.02 kb/s, Avg QP:26.24

 

[biostud]

Win10, 5820K@4.4Ghz, DDR4-2133

encoded 1497 frames in 138.26s (10.83 fps), 4036.02 kb/s, Avg QP:26.24

 

[Brunnis]

Core i7-6700K @ stock, DDR4-2666, Windows 10

encoded 1497 frames in 185.77s (8.06 fps), 4036.02 kb/s, Avg QP:26.24

 

[cbuchach]

x265 [info]: consecutive B-frames: 26.4% 9.1% 11.5% 42.3% 10.7%
encoded 1497 frames in 155.57s (9.62 fps), 4036.02 kb/s, Avg QP:26.24

i7 7700K at 5 GHz 3200MHz DDR4

 

[MarkLuvsCS]

encoded 1497 frames in 136.64s (10.96 fps), 4036.02 kb/s, Avg QP:26.24

i7-6850k @ 4.3ghz, 3200mhz ddr4 quad-channel, windows 10

 

[Huberopa]

encoded 1497 frames in 616,71s (2,43 fps), 4036,02 kb/s, Avg QP:26,24

AMD A8-5600K 3.60GHz (stock), 16GB DDR3@1866. Ubuntu 16.04.1LTS

 

[BonzaiDuck]

I came back to see how my own results fit with the rank-ordering. I have a couple questions to pose.

First -- would the benchmark benefit more from processors sporting more cores (six, eight, twelve) as opposed to less (for instance -- four)?

Second -- would SLI or CF -- two or more -- be tie-breaker?

 

[lopri]

Holy kangaroo 1045T sucks at this. I expected to see somewhere along A10-7850K-ish number, but instead it turns out slower than Larry's Pentium G4600. And It is like below 20% speed of the dual Xeon E5-2670 setup. It was a dirty run done while I was reading emails but it took so long that I did not bother to do a clean run.

encoded 1497 frames in 695.45s (2.15 fps), 4036.02 kb/s, Avg QP:26.24

1045T @3.9 GHz (13x300)
8 GB of DDR3-800 / 7-8-7-20
Windows 7 x64

 

[SPBHM]

that's the lack of SSE 4.1 as mentioned earlier, a 3GHz Core 2 Quad (45nm) should be able to beat that result

regarding the results table, I see some people are declaring base clock when the CPU actually runs at boost clock, so it's kind of confusing.

 

[Teizo]

HandBrake 1.0.2 (2017012200) - 64bit
OS: Microsoft Windows NT 10.0.14393.0 - 64bit
CPU: Intel(R) Core(TM) i7-7700K CPU @ 4.20GHz
Ram: 16341 MB,
GPU Information:
NVIDIA GeForce GTX 1080 - 21.21.13.7857
Screen: 2560x1440

i7 7700k @ 5ghz
DRAM @ 3200mhz

encoded 1497 frames in 155.38s (9.63 fps), 4036.02 kb/s, Avg QP:26.24

 

[ElFenix]

i7 7700k at 5ghz DDR4-3000 Windows 10. It throttles down to 4.8 sometime. I guess it's the AVX code? Too many things are still set at auto. It did hit 88c so lowering the clock for AVX is a good thing.

encoded 1497 frames in 158.29s (9.46 fps), 4036.02 kb/s, Avg QP:26.24

will need to update from this post

 

[CentroX]

encoded 1497 frames in 272.15s (5.50 fps), 4036.02 kb/s, Avg QP:26.24

 

[Jan Olšan]

Here's some Piledriver/Richland.

HandBrake 1.0.1 (2016122900) 64bit, W10 64bit

AMD A10-6800K 4.1-4.4 GHz @ stock clocks/no OC (I think it boosts to between 4.20 and 4.30 GHz for the encode load)
Ram: 8 GB (512 MB shared for iGPU), dual DDR3 1600 MHz CL11
FM2 / Gigabyte F2A85X-UP4

encoded 1497 frames in 566.78s (2.64 fps), 4036.02 kb/s, Avg QP:26.24

 

[ElFenix]

encoded 1497 frames in 272.15s (5.50 fps), 4036.02 kb/s, Avg QP:26.24

need processor and other info

also, i'll need to do some tests to see if 1.02 changes anything

 

[Teizo]

also, i'll need to do some tests to see if 1.02 changes anything

Oh, my bad. I used the new version. I just followed the link and downloaded the one I saw for Windows and missed the archive link.

 

[mrpiggy]

HandBrake 1.0.1 (2016122900) - 64bit
OS: Microsoft Windows NT 10.0.14393.0 - 64bit
CPU: Intel(R) Xeon(R) CPU E5-2690 v3 @ 2.60GHz
Ram: 65456 MB,
GPU Information:
NVIDIA GeForce GTX 980 - 21.21.13.7866
Screen: 2560x1440
RAM is DDR4 2133 (Xeon V3's don't support faster)

encoded 1497 frames in 98.08s (15.26 fps), 4036.02 kb/s, Avg QP:26.24

All 12 cores stayed between 3.4 and 3.6 Ghz while running.

 

[CentroX]

need processor and other info

also, i'll need to do some tests to see if 1.02 changes anything

4820k @ 4.3 ghz
16GB ddr3 1833
Win 10

 

[ElFenix]

someone out there has a ryzen processor they want to test

 

[AMDisTheBEST]

I can't afford Ryzen to test it out with.

 

[Crono]

someone out there has a ryzen processor they want to test

encoded 1497 frames in 132.90s (11.26 fps), 4036.02 kb/s, Avg QP:26.24

Processor: AMD Ryzen 7 1700X
Processor Speed: 3.6GHz
RAM: DDR4-2666
Operating System: Windows 10