DVD Rebuilder Pro + HCenc + i7

[soneca]

Hi jdobbs,

The use of multiple threads with HC has always been great using the i7 920, but now with the i7 980x greatly diminished.

So I decided to disable Hyper-threading and performance improved considerably.

Any idea how to optimize performance?

With ht:o
[11:57:22] Phase I, PREPARATION started.
- DVD-RB v1.28.2
- AVISYNTH 2.5.8.0
- HC v0.25.0.0 encoder selected
- Source: O_BOM_FILHO_A_CASA_TORNA
- VTS_01: 3.506.180 sectors.
-- Scanning and writing .D2V & .AVS files
-- Processed 163.812 frames.
-- Building .AVS and .ECL files
- Reduction Level for DVD-5: 59,5%
- Overall Bitrate : 5.586/4.469Kbs
- Space for Video : 3.727.152KB
- Redistributing using Base_Q: 23
- HIGH/LOW/TYPICAL Bitrates: 5.836/786/4.469 Kbs
[12:00:21] Phase I, PREPARATION completed in 3 minutes.
[12:00:21] Phase II ENCODING started
- Creating M2V for VTS_01 segment 0
- Creating M2V for VTS_01 segment 1
- Creating M2V for VTS_01 segment 2
- Creating M2V for VTS_01 segment 3
- Creating M2V for VTS_01 segment 4
- Creating M2V for VTS_01 segment 5
- Creating M2V for VTS_01 segment 6
- Creating M2V for VTS_01 segment 7
- Creating M2V for VTS_01 segment 8
- Creating M2V for VTS_01 segment 9
- Creating M2V for VTS_01 segment 10
- Creating M2V for VTS_01 segment 11
- Creating M2V for VTS_01 segment 12
- Creating M2V for VTS_01 segment 13
- Creating M2V for VTS_01 segment 14
- Creating M2V for VTS_01 segment 15
- Creating M2V for VTS_01 segment 16
- Creating M2V for VTS_01 segment 17
- Creating M2V for VTS_01 segment 18
- Creating M2V for VTS_01 segment 19
- Creating M2V for VTS_01 segment 20
- Extracting STILLS for VTS_01 segment 21
[12:14:09] Phase II ENCODING completed in 14 minutes.

No ht:)
[18:39:12] Phase I, PREPARATION started.
- DVD-RB v1.28.2
- AVISYNTH 2.5.8.0
- HC v0.25.0.0 encoder selected
- Source: O_BOM_FILHO_A_CASA_TORNA
- VTS_01: 3.506.180 sectors.
-- Scanning and writing .D2V & .AVS files
-- Processed 163.812 frames.
-- Building .AVS and .ECL files
- Reduction Level for DVD-5: 59,5%
- Overall Bitrate : 5.586/4.469Kbs
- Space for Video : 3.727.152KB
- Redistributing using Base_Q: 23
- HIGH/LOW/TYPICAL Bitrates: 5.836/786/4.469 Kbs
[18:42:04] Phase I, PREPARATION completed in 3 minutes.
[18:42:04] Phase II ENCODING started
- Creating M2V for VTS_01 segment 0
- Creating M2V for VTS_01 segment 1
- Creating M2V for VTS_01 segment 2
- Creating M2V for VTS_01 segment 3
- Creating M2V for VTS_01 segment 4
- Creating M2V for VTS_01 segment 5
- Creating M2V for VTS_01 segment 6
- Creating M2V for VTS_01 segment 7
- Creating M2V for VTS_01 segment 8
- Creating M2V for VTS_01 segment 9
- Creating M2V for VTS_01 segment 10
- Creating M2V for VTS_01 segment 11
- Creating M2V for VTS_01 segment 12
- Creating M2V for VTS_01 segment 13
- Creating M2V for VTS_01 segment 14
- Creating M2V for VTS_01 segment 15
- Creating M2V for VTS_01 segment 16
- Creating M2V for VTS_01 segment 17
- Creating M2V for VTS_01 segment 18
- Creating M2V for VTS_01 segment 19
- Creating M2V for VTS_01 segment 20
- Extracting STILLS for VTS_01 segment 21
[18:53:34] Phase II ENCODING completed in 11 minutes.

[Groucho2004]

Processed 163.812 frames.
ENCODING completed in 11 minutes.

Right, 11 minutes for 164,000 frames is still too slow for you.

How many instances of HC are you running? Multithreaded or not? Why don't you experiment yourself?

[soneca]

12 instances with ht enabled and 6 instances with disabled ht.
Is very fast but do not understand is the decrease in performance with HT enabled because it never occurred using HC.
Maybe Hank has to optimize the HC for a larger number of threads...

[soneca]

This article (http://www.tomshardware.com/reviews/hyper-threading-core-i7-980x,2584.html) is helping me to better understand the Hyper-Threading on the i7 980x.

[soneca]

After some tests I noticed that very few applications take advantage of ht and others to lose performance.
The x264 is the most benefits from ht but with little difference and a considerable increase in consumption and cpu temperature.
So Hyper-Threading will be disabled from now.

[SLOVEHEART]

Hi jdobbs,

[11:57:22] Phase I, PREPARATION started.
- DVD-RB v1.28.2
- AVISYNTH 2.5.8.0
- HC v0.25.0.0 encoder selected


Just a quick side question - I notice your using the latest HC encoder version 25. I've also run DVD-RB Pro w/ HC 25 renamed as HCbatch.exe - and in the status window it lists as HC v0.23... - how are you getting the program to list the correct version??

BTW - I use AVISYNTH 2.6.0 - this may be a performance gain for you??? Your encoding times seem quite good to me - I run a Core i7 950 cpu (I know - stupid waste of money) at 3.616 GHz and my encoding takes about 16-17 minutes using 8 threads (HT enabled)...

Cheers...

[soneca]

Nothing, ever renamed to HCbatch to replace the old file and he was identified correctly by the DVD Rebuilder.
Actually version 2.6.0 is pretty stable, I'll test it soon.
Never had problems related to HT, but with this new processor that gain is questionable ...

[jdobbs]

Just a quick side question - I notice your using the latest HC encoder version 25. I've also run DVD-RB Pro w/ HC 25 renamed as HCbatch.exe - and in the status window it lists as HC v0.23... - how are you getting the program to list the correct version??

BTW - I use AVISYNTH 2.6.0 - this may be a performance gain for you??? Your encoding times seem quite good to me - I run a Core i7 950 cpu (I know - stupid waste of money) at 3.616 GHz and my encoding takes about 16-17 minutes using 8 threads (HT enabled)...

Cheers... DVD-RB pulls the version from the program itself. If yours is saying v0.23 -- you're probably actually running v0.23.

[hank315]

@soneca
AFAIK a hyperthreaded CPU shares the MMX/SSEx registers.
Because HCenc uses MMX/SSEx instructions a lot it's best to set it to 6 instances for a I7 980x.
Using 12 instances also means you need *a lot* of memory.

[soneca]

Hank,

Thanks for clarifying.
With 12 instances the memory consumption is really huge and the performance drops almost 30%.
But the speed of HC remains unbeatable.;)

[SLOVEHEART]

DVD-RB pulls the version from the program itself. If yours is saying v0.23 -- you're probably actually running v0.23.

Well- you're right; I was sure I saw ver. 25 listed as 23, I guess I must have been blind :) - I just re did the switch; the version is labeled correctly!!

BTW - ver 25 doesn't hang on initiation like the previous beta did. Encoding times are the same for the most part - not tested fully...

Cheers...