Android 2.2 - Froyo

[Tanknspank]

I agree, I would really like to see that new launcher. But I suppose I would still be satisfied if we just get that speed boost from the new just in time compilation addition.

We will notice a very nice improvement from JIT now that it's officially supported by Google and we will get software that can take advantage of it.

any idea why the Droid is not seeing the same performance gains on Linpack as the N1. This is only ~2x not 4-5x as seen on the N1.

We have no proof that this is even official. So don't get disappointed just yet.

Though, I have no real idea as to why it is so small as compared to the N1, if this does turn out to be legit.

 

[hookbill]

Guys.....and Girls....whomever....Family Friendly language. You're making the censor blow up.

 

[jbwiden]

We will notice a very nice improvement from JIT now that it's officially supported by Google and we will get software that can take advantage of it.

any idea why the Droid is not seeing the same performance gains on Linpack as the N1. This is only ~2x not 4-5x as seen on the N1.

We have no proof that this is even official. So don't get disappointed just yet.

Though, I have no real idea as to why it is so small as compared to the N1, if this does turn out to be legit.

True... For me, every post stemming from these latest reports includes an implicit assumption that the report is true and is a release version similar to the one which will be pushed to the Droid.

 

[nphil]

Well the MFLOPS is almost twice as what you'd see on a stock 600 Mhz Droid. The N1's Snapdragon is clocked at 1Ghz stock and is slightly different from the OMAP's implementation of the cortex A8.

I'm interested in overclocked numbers, or even if the phone would be stable at higher speeds.

 

[natty_lite]

10 MFLOPS is still faster than the 8.5 I average on my while Droid while OC'ed to 1GHZ. You mean to tell me I can get performance like that on a stock kernel? SOLD!.

This update will probably be enough to keep my phone interesting enough till Verizon moves forward with their ever elusive "Shadow"

 

[Shadez]

I personally dont believe so much hype about all the features would be out their or even mentioned if we werent getting them. I mean come on, why would they even bother mentioning it at the I/O ??? Seriously.. its like YES 2.2 CAN DO ALL THIS... but you cant use it. Everyone relax till its released ..

Don't make me turn this car around, I will ya know...

 

[JhankG]

Didn't the Nexus with 2.2 get around 39 mflops? Maybe an OCed Droid with 2.2 will get something a bit closer to that number?

 

[boostdscoob]

with JIT at stock clock, you can get around 6-8mflops, so yes, i dont believe this is a release version, or even a test version. that or the kernel isnt applied yet.

 

[Darkseider]

The Linpack benchmark scales linearly with the clockspeed of a device.

On a Moto Droid @ 800 Mhz the avg. score is 6.5 MFlops which breaks down to roughly .815 MFlops per 100 Mhz. Therefore a 1 Ghz Droid scores around 8.15 MFlops and a 1.2 Ghz Droid around 9.78 MFlops. With that in mind and the claimed 4.5x performance boost JIT gives to Linpack, which is a synthetic benchmark, a 1Ghz Droid should score approx. 36.675 on Linpack. Which is directly in line with the Nexus One @ 1Ghz.

The linpack while a nice little test is nothing more than a pissing contest and doesn't really show any true performance benchmarks.

 

[P3Droid]

with JIT at stock clock, you can get around 6-8mflops, so yes, i dont believe this is a release version, or even a test version. that or the kernel isnt applied yet.

boostdscoob, wrong on all accounts, lol. but nice speculation.

 

[JhankG]

The Linpack benchmark scales linearly with the clockspeed of a device.

On a Moto Droid @ 800 Mhz the avg. score is 6.5 MFlops which breaks down to roughly .815 MFlops per 100 Mhz. Therefore a 1 Ghz Droid scores around 8.15 MFlops and a 1.2 Ghz Droid around 9.78 MFlops. With that in mind and the claimed 4.5x performance boost JIT gives to Linpack, which is a synthetic benchmark, a 1Ghz Droid should score approx. 36.675 on Linpack. Which is directly in line with the Nexus One @ 1Ghz.

The linpack while a nice little test is nothing more than a pissing contest and doesn't really show any true performance benchmarks.

I know it doesn't really translate into faster speeds, but seriously, I can pee further than you guys can.

 

[Shadez]

I bet Steve Jobbs is the one who started circulating these pics...

 

[boostdscoob]

right, then lets see some proof, because i just clicked a 5.897mflop score at 250min/600max.
so 2mflops is definitely not the increase we should see.

 

[boostdscoob]

how about a short bench test score P3?

 

[jntdroid]

I bet Steve Jobbs is the one who started circulating these pics...

agreed - he's causing dissension in the ranks!