Subject | Re: Google's answer to the iPad3 has just arrived... |
From | KDT |
Date | 11/02/2012 22:10 (11/02/2012 14:10) |
Message-ID | <560ca837-62a1-4def-b51b-09461cb98af5@googlegroups.com> |
Client | |
Newsgroups | comp.sys.mac.advocacy |
Follows | ed |
Followups | Flint (2h & 59m) > KDT ed (3h & 38m) > KDT FDK (4h & 21m) > KDT |
edthe mali 400, which is known to be an excellent gpu. saying that we'll wa= it for benchmarks to see if it has good peformance is like saying we'll nee= d to wait for benchmarks to see if a new quad core 3ghz xeon mac pro has be= tter performance than an old core duo mac.
On Thursday, November 1, 2012 4:12:53 AM UTC-7, KDT wrote: =20On Thursday, November 1, 2012 12:19:43 AM UTC-4, ed wrote:=20=20We will see when the benchmarks come out....the t604 is a better class, higher clocked fpu, with more cores, than=
=20-------------=20
r benched slower than an iPad even with a faster CPU.=20=20Because it's not like a flagship Android tablet (i.e. the Xoom) has eve=
=20 =20 =20 almost irrelevant, and largely incorrect. you must be talking xoom vs og =ipad if you're talking a slower cpu in the ipad. and i've noticed you've m= oved from the gpu to the cpu. jeebus, stick to one thing at a time... but= what benchmark would you like to compare? here's a couple to get you star= ted:
=20 http://www.glbenchmark.com/compare.jsp?benchmark=3Dglpro25&showhide=3Dtru=e&certified_only=3D1&D1=3DMotorola%20Xoom&D2=3DApple%20iPad=20
=20 http://jeftek.com/1942/motorola-xoom-sunspider-results/ =20 =20 =20 =20 =20 but look, we know the last gen mali performs very well on android. this =thing is going to be better in almost all respects. i can't believe, based= on just that, you won't admit this thing is going to be a performer. ;D
=20 =20 =20 =20 =20the graphics performance?Without actually seeing the display, or knowing anything about =
=20fastest chips....=20=20=20we know the mali t604 is fast...=20=20And we also know that the Dalvik VM can suck performance out of the=
=20dia-tegra-2-chip=20=20=20not really.=20=20http://techztalk.com/techwebsite/03-19-11-apple-a5-processor-smokes-nvi=
=20marks suck as a general purpose benchmark, but ok. ;D what you show doesn= 't even begin to address your point though- how much of the deltas above ar= e you attibuting to dalvik, and how much to hardware differences? i.e. on = the glbench, we know the results of textures would be mostly hardware- it's= well known the tegra 2 sucks there...=20=20And yes Linpack does use Dalvik...=20=20=20http://www.legitreviews.com/article/1562/4/=20 =20 =20 linpack kind of sucks as a general purpose benchmark... well, most bench=
=20 =20 =20ust as fast as native" would you? I thought those people died out over a = decade ago.You wouldn't happen to be one of those people who think Java can run "j=
=20 =20 =20 certainly not for general purposes, and certainly not swing. some things=, sure (i've done the benchmarks).