[Xsever]

coppolla, on 21 April 2012 - 07:52 AM, said:

CyanogenMod includ thé New driver in nightly build

Envoyé depuis mon SGH-T989 avec Tapatalk

Merci pour l'info!

Sent from my ICS Touchpad using Tapatalk

[Dorregaray]

Xsever, on 21 April 2012 - 06:08 AM, said:

Sent from my cm_tenderloin using Tapatalk

The one with "new"at the end.

I wonder if the results on the ones without "new" will be any different

[travisross69]

Check the gethub fornthe updates on the CM nightly. It doesn't say that the new drivers have been merged yet, but the unofficial nightly they have. You can check the official nightly discussion thread in development to follow what is and isn't merged. I run the officials, but I reflash the drivers each time I update the rom.
Here is a link to the discussion http://rootzwiki.com...ild-discussion/

Edited by travisross69, 21 April 2012 - 12:06 PM.

[O.a.T.]

Xsever, on 20 April 2012 - 06:16 AM, said:

New Driver results with AnTuTu: 2D: 153 3D: 1177

with Quadrant: 2D: 324 3D: 1790

Are your results, both on the old and new, from a single run on each or do these represent some kind of average over multiple runs?

[Xsever]

The numbers that say 'old' are for the first Ardeno drivers that surfaced and those that say 'new' are for the newest ardeno drivers.

[gamer765]

Yes, I think we got that already. What he was asking was, did you run the benchmark once and give us that result, or did you run it several times and average the results?

[Xsever]

gamer765, on 21 April 2012 - 09:19 PM, said:

Yes, I think we got that already. What he was asking was, did you run the benchmark once and give us that result, or did you run it several times and average the results?

I was answering Dorregaray.

I ran the benchmarks once. No averaging. With cpu governer set to performance.

[Dorregaray]

The 'new' drivers are obtained from other device while the 'old' are released by the Qualcomm so the new ones may be optimised for other device type not exactly the TP so maybe the 'old' ones are better for TP?
And what are he benchmarks running on stock drivers from current cm9 nightlies?

[O.a.T.]

Xsever, on 21 April 2012 - 09:45 PM, said:

I was answering Dorregaray.

I ran the benchmarks once. No averaging. With cpu governer set to performance.

Hard to draw any conclusions then as I can see as much variance and more from one run to another within a given benchmark on the same setup. To make it meaningful you really need to run multiple times and pick some way to normalize.

[travisross69]

Bump, to keep it up!