View Poll Results: For better fairness, should I...
...continue to benchmark with T132 20 18.52%
...use the kernel supplied with the ROM instead 34 31.48%
...do it both ways 54 50.00%
Voters: 108. You may not vote on this poll

Thread Closed

[REF] All ICS ROMs Benchmarked

bedalus
7th April 2012, 01:53 PM   |  #791  
Guest
Thanks Meter: 0
 
n/a posts
UPDATE: User Experience benchmark is up.

The words shown occur more than average (where the baseline is generated from words extracted from ALL threads).

If a word is occurring more than average, I see this as significant. The words are ranked in significance from #1 to #10 according to how far above the baseline they are.

I'm not going to try and interpret whether each word has a positive or negative implication, I leave that up to you.

Attached is the complete data. Also see here for how I developed the software to help me do all this.
Attached Files
File Type: xlsx ALLROM_word_baseline.xlsx - [Click for QR Code] (69.3 KB, 40 views)
Last edited by bedalus; 7th April 2012 at 03:54 PM.
The Following 4 Users Say Thank You to For This Useful Post: [ View ]
bedalus
7th April 2012, 01:56 PM   |  #792  
Guest
Thanks Meter: 0
 
n/a posts
Quote:
Originally Posted by Oodie

One thing i can get from colomn f is , that no kernel can match T132 ?

( it's obvious that you try the stock kernel with 1000 & T132 on 1320 . It's the reason right ? So does the colomn f reasonable regurding other kernels which tested with 1000mhz ?)

Okay, column F now shows the boost gained from flashing T132.

The average boost is 18%
7th April 2012, 01:57 PM   |  #793  
tchaari's Avatar
Senior Member
Thanks Meter: 1,357
 
1,124 posts
Join Date:Joined: Dec 2011
Quote:
Originally Posted by Oodie

One thing i can get from colomn f is , that no kernel can match T132 ?

( it's obvious that you try the stock kernel with 1000 & T132 on 1320 . It's the reason right ? So does the colomn f reasonable regurding other kernels which tested with 1000mhz ?)

With colomn F, we can have an idea about how much the stock kernel is far from T132 even if they are not equally clocked
7th April 2012, 02:14 PM   |  #794  
Oodie's Avatar
Recognized Developer
Flag Kuliyapitiya
Thanks Meter: 1,966
 
1,635 posts
Join Date:Joined: May 2011
Donate to Me
More
Quote:
Originally Posted by bedalus

Okay, column F now shows the boost gained from flashing T132.

The average boost is 18%

That's better

Now i know you hired ninja assassins to kill me . One just tried to kill me , got a slap from my wife and went away he dropped a cheque given from a man named " bedalus " .
bedalus
7th April 2012, 02:19 PM   |  #795  
Guest
Thanks Meter: 0
 
n/a posts
Quote:
Originally Posted by Oodie

That's better

Now i know you hired ninja assassins to kill me . One just tried to kill me , got a slap from my wife and went away he dropped a cheque given from a man named " bedalus " .

Tear that cheque up. He can't get paid if he doesn't do the job.

Now for the correct smiley. Umm... also

Mainly
The Following User Says Thank You to For This Useful Post: [ View ]
7th April 2012, 03:07 PM   |  #796  
glennkaonang's Avatar
Senior Member
Flag Surabaya
Thanks Meter: 194
 
509 posts
Join Date:Joined: Feb 2012
More
I still don't get the idea of column F.
Statistics has always been a subject I couldn't catch up with.

So, in Column F, is it the higher the better or otherwise?

Sent from my Nexus S using xda premium
7th April 2012, 04:02 PM   |  #797  
Oodie's Avatar
Recognized Developer
Flag Kuliyapitiya
Thanks Meter: 1,966
 
1,635 posts
Join Date:Joined: May 2011
Donate to Me
More
lol. just read the user experience column . & it's just ... "lol"
bedalus
7th April 2012, 04:06 PM   |  #798  
Guest
Thanks Meter: 0
 
n/a posts
Quote:
Originally Posted by glennkaonang

I still don't get the idea of column F.
Statistics has always been a subject I couldn't catch up with.

It shows how much of a boost it got from flashing the T132 kernel.

Compare Column F for AOSP+(n.5) and Peter Alfonso's Bugless Beast Nightly.

AOSP+ 26%
Bugless Beast 13%

AOSP+
When I flashed T132 on AOSP+ the result went up 26%
This is more than the average of 18%
You get a bigger gain than expected by flashing T132
So... maybe it is a good idea to replace the kernel on this ROM.

Bugless Beast
When I flashed T132 on BB the result went up 13%
This is less than the average of 18%
You get a smaller gain than expected by flashing T132
So... maybe it is a good idea to keep the default kernel.
The Following 5 Users Say Thank You to For This Useful Post: [ View ]
7th April 2012, 05:03 PM   |  #799  
glennkaonang's Avatar
Senior Member
Flag Surabaya
Thanks Meter: 194
 
509 posts
Join Date:Joined: Feb 2012
More
That's it, I got it now.
Those you bolded really explains much.
I'm that kind of guy who understands words better than numbers

Thanks a lot, bedalus.
7th April 2012, 08:15 PM   |  #800  
AeroEchelon's Avatar
Senior Member
Flag Toronto, ON
Thanks Meter: 67
 
367 posts
Join Date:Joined: Nov 2011
More
Sweet so what I gather from this is Pete's ROM has the best scores on benchmarking if all the ROMs were pitted against each other using the supplied kernel?

(Also, the User Experience column says that "kernel" is the most said word in his thread. And probably for good reason.)

Thread Closed Subscribe to Thread
Previous Thread Next Thread
Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes


Top Threads in Nexus S General by ThreadRank