Benchmark your computer with Handbrake 1.01 and x265

[SPBHM]

AVX/2 causes the difference.

the E5420 doesn't support AVX and it got the same as the AVX CPUs, it's something else.

 

[ElFenix]

with a "Core 2 Duo" Pentium(T4400):
encoded 1497 frames in 2798.17s (0.53 fps), 3978.85 kb/s, Avg QP:26.14

if you can disable 2 cores of a quad core to test if it changes the bitrate it could be interesting

AVX/2 causes the difference.

i just ran it on an i3 with hyperthreading off, it's actually a dual core thing.

 

[lopri]

encoded 1497 frames in 133.08s (11.25 fps), 4036.02 kb/s, Avg QP:26.24

Dual Xeon E5-2670s (2.6/3.3) that I have been testing out. No HT, dual-channel DDR3-1333. Runs at 3.0 GHz throughout so at least on this system AVX is not at work.

 

[BonzaiDuck]

OK . . . HEre it is -- my 6700K Skylake sig system: i7-6700K @ 4.7 Ghz; DDR4-3200 14-14-14 @ XMP spec; Z170 in Sabertooth Z170 S. Currently booted to the Win7-64 OS. Hyperthreading enabled.

I'd never tried running this before, although I thought it was a component of ROG RealBench, which I've used in stress-testing. My CPU temperatures went up to about 60C +/- all cores.

Here's the output line:

encoded 1497 frames in 168.77s (8.87 fps), 4036.02 kb/s, Avg QP:26.24

Hmm . . . falls short of bbhaag's i7-6700K @ 4.6. What does that mean? I've got a single GTX 1070 OC'd to ~2,038 core and 8,904 Mhz memory.

 

[pm]

Xeon E5-2699v4@2.2GHz, DDR4-2400 (Broadwell server, 22 core, 55MB L2)
encoded 1497 frames in 85.59s (17.49 fps), 4058.76 kb/s, Avg QP:26.32

Surprisingly close to The Stilt's number considering the 4 extra cores, slightly faster memory and the 10MB more cache.

Edit: clean reboot knocked 1s off of my score... no idea why except that Windows likes it's reboots. Updated score.

 

[superstition]

ElFenix, small correction to your leaderboard. The slower of the two Macbook Pros was running macOS 10.12.2 not 10.9.5.

Here is a new score for your list:

263.06s (5.69 fps), 4036.02 kb/s, Avg QP:26.24

8320E, 4.4 GHz, Win 8.1, 2133 9-11-10-30-1T

 

[rancherlee]

FX-8320 @ 4.4 ghz 1866 ddr3 9-10-9-27-1T windows 10

encoded 1497 frames in 270.87s (5.53 fps), 4036.02 kb/s, Avg QP:26.24


I ran it a few times and it varies about 10 seconds, that one is off a fresh boot.

 

[Kallogan]

Mobile core i5 6300hq 2.8ghz on 4 cores and 8gb ddr4 2133MHz

320.98s

Not that bad i guess ddr4 helps in encoding

Power consumption is ridiculously low. 45W at the wall, knowning that this gaming laptop idles at 21W.
cpu is undervolted at -0.115V in xtu. 43W if i go at -0.130.

 

[The Stilt]

Xeon E5-2699v4@2.2GHz, DDR4-2400 (Broadwell server, 22 core, 55MB L2)
encoded 1497 frames in 85.59s (17.49 fps), 4058.76 kb/s, Avg QP:26.32

Surprisingly close to The Stilt's number considering the 4 extra cores, slightly faster memory and the 10MB more cache.

Edit: clean reboot knocked 1s off of my score... no idea why except that Windows likes it's reboots. Updated score.

X265 scales extremely poorly above ~ 10 cores at 1080P due it's large CU size. Technically it should scale up to 17 threads at 1080P (or 34 threads at 2160P), however based on the measurements I've made the scaling starts to decline significatly already after 4 cores.

Cores / FPS / Scaling @ 1080P

1 = 2.26fps
2 = 4.57fps (102.21%)
4 = 8.58fps (87.74%)
8 = 15.57fps (81.46%)
16 = 24.95fps (60.24%)

 

[PG]

Xeon E5-2630 V4 ES QK3G 10 cores, 20 threads, 2.2Ghz base, 3.1Ghz turbo
Asrock X99 Extreme4/3.1
32 GB (4 x 8GB) Crucial Ballistix Sport LT DDR4-2400 at 2133Mhz 16-16-16-39
Samsung 850 Evo 50GB

encoded 1497 frames in 147.32s (10.16 fps), 4036.02 kb/s, Avg QP:26.24

 

[Hail The Brain Slug]

encoded 1497 frames in 98.92s (15.13 fps), 4036.02 kb/s, Avg QP:26.24

Rebooted from messy work install into clean gaming install, cut off 0.8 seconds.
encoded 1497 frames in 98.12s (15.26 fps), 4036.02 kb/s, Avg QP:26.24

5960x @ 4.5 core/4.375 cache, DDR4 3000 14-16-16-32 Windows 10

Shouldn't the leaderboard list the C/T count and frequency for The Stilt's Haswell EP HCC processor since it's not apparent or even well defined?

 

[lopri]

Blame Intel for confusing the consumers even the enthusiasts.

 

[Hail The Brain Slug]

Blame Intel for confusing the consumers even the enthusiasts.

Its not Intel's fault he listed the type of CPU and not the exact model (providing its not some special EP model without a model? Do they even make those?). The HCC core is used for a variety of models ranging from 14-18 cores.

He did list the important info (core count and frequency) but that didn't make it into the OP leaderboard.

 

[Hail The Brain Slug]

Oops, double post.

 

[asendra]

Win 10 // i7-4770k@4.2Ghz DDR3-1600

encoded 1497 frames in 204.90s (7.31 fps), 4036.02 kb/s, Avg QP:26.24

PD: Dud OC chip, I have a heat wall with 1.25v but haven´t had the time and need to delid it. Also, I've always regretted not buying faster ram (ddr3 2133) when I built this. I would love to see how much differece It makes in haswell.


Still, 25% slower than a 10% higher clocked Skylake with much faster ram. It really shows how little there's to gain by upgrading right now.

 

[Hail The Brain Slug]

Win 10 // i7-4770k@4.2Ghz DDR3-1600

encoded 1497 frames in 204.90s (7.31 fps), 4036.02 kb/s, Avg QP:26.24

PD: Dud OC chip, I have a heat wall with 1.25v but haven´t had the time and need to delid it. Also, I've always regretted not buying faster ram (ddr3 2133) when I built this. I would love to see how much differece It makes in haswell.


Still, 25% slower than a 10% higher clocked Skylake with much faster ram. It really shows how little there's to gain by upgrading right now.

My 4770k would barely do 4.0 before I delidded it. Afterward I was able to get 4.3 stable.

Do I get to do two entries? if so,

4770k 4.3 core/4.1 cache DDR3 2400 10-12-12-32 Windows 10
encoded 1497 frames in 193.74s (7.73 fps), 4036.02 kb/s, Avg QP:26.24

193.74 vs my 5960x at 4.5 GHz at 98.12 shows some bad scaling from 4c/8t to 8c/16t

 

[pm]

My 4770k would barely do 4.0 before I delidded it. Afterward I was able to get 4.3 stable.

Do I get to do two entries? if so,

4770k 4.3 core/4.1 cache DDR3 2400 10-12-12-32 Windows 10
encoded 1497 frames in 193.74s (7.73 fps), 4036.02 kb/s, Avg QP:26.24

193.74 vs my 5960x at 4.5 GHz at 98.12 shows some bad scaling from 4c/8t to 8c/16t

Yeah, my previous computer before my current one was a 4770K as well and I had a similar experience. I t could run at 4.1GHz before I delidded (at close to stock voltage) and got about 200MHz by delidding it.

 

[Makaveli]

if someone could collect the results and put them in an excel graph that would be nice

 

[bbhaag]

OK . . . HEre it is -- my 6700K Skylake sig system: i7-6700K @ 4.7 Ghz; DDR4-3200 14-14-14 @ XMP spec; Z170 in Sabertooth Z170 S. Currently booted to the Win7-64 OS. Hyperthreading enabled.

I'd never tried running this before, although I thought it was a component of ROG RealBench, which I've used in stress-testing. My CPU temperatures went up to about 60C +/- all cores.

Here's the output line:

encoded 1497 frames in 168.77s (8.87 fps), 4036.02 kb/s, Avg QP:26.24

Hmm . . . falls short of bbhaag's i7-6700K @ 4.6. What does that mean? I've got a single GTX 1070 OC'd to ~2,038 core and 8,904 Mhz memory.

Could be I'm running Windows10 pro. I also have two ssds so I set it up too read from one drive and write too the other. That's all I can think of as too why the scores would be different. My gpu is a Rx480 4gb and my mb is a Asrock Z170M Extreme4 so I doubt that made a difference.

I just ran the test again after a restart and managed to knock a few more tics off my first time.
encoded 1497 frames in 163.02s (9.18 fps), 4036.02 kb/s, Avg QP:26.24

 

[Conroe]

Shaved a few seconds off with the only overclock I can muster out of this locked CPU.
Window 10, i5 6500 with a 102.9 mhz BCLK (3396 mhz four core turbo) and DDR4 3018 CL16, second run.
encoded 1497 frames in 269.75s (5.55 fps), 4036.02 kb/s, Avg QP:26.24

 

[crashtech]

CPU: Intel(R) Core(TM) i3-4370 CPU @ 3.80GHz
encoded 1497 frames in 443.98s (3.37 fps), 4036.02 kb/s, Avg QP:26.24

 

[Edgemeal]

i7-3770 @ 4.0GHz (100x40) 16GB DDR3 8-8-8-24-2T Windows 7.
encoded 1497 frames in 281.67s (5.31 fps), 4036.02 kb/s, Avg QP:26.24

I gather some scores and try to make chart.

 

[Hail The Brain Slug]

i7-3770 @ 4.0GHz (100x40) 16GB DDR3 8-8-8-24-2T Windows 7.
encoded 1497 frames in 281.67s (5.31 fps), 4036.02 kb/s, Avg QP:26.24

I gather some scores and try to make chart.

Awww, missed my 4770k score. Oh well.

Looks really good, though. :beer:

 

[superstition]

i7-3770 @ 4.0GHz (100x40) 16GB DDR3 8-8-8-24-2T Windows 7.
encoded 1497 frames in 281.67s (5.31 fps), 4036.02 kb/s, Avg QP:26.24

I gather some scores and try to make chart.

Nice job, but you're missing some scores. My latest is missing.

263.06s (5.69 fps), 4036.02 kb/s, Avg QP:26.24, 8320E, @4.4 GHz, Win 8.1, 2133 9-11-10-30-1T

Personally, I would list construction core processors as modules rather than cores for greater accuracy. So, 4M/8T for full Piledriver.

 

[crashtech]

(redacted)

My apologies to superstition and the OP.