RgbenchMM

Man i got a 8 cores ( 4 x 2.1GHz + 4 x 1.5GHz ) 8 cores of Samsung Exynos Octa 7420 2.10 GHz help plzzz !!!

4617 with quad on my lg g2. 5400 on my gs6 on quad. But that's only half my cores. My a8x only gets a 2700?

2.2 Gflops on 2 cores, it's like CRAY-2

needs to be updated

Perfect for device/rom comparisons. I'll use it to update my results about snapdragon 400 behavior (previously done with an old outdated app)

2.2 Gflops on 2 cores, it's like CRAY-2

Desperately needs an update. Just for kicks I ran it on a store model Galaxy S6 yesterday and it reported over 5000.

4617 with quad on my lg g2. 5400 on my gs6 on quad. But that's only half my cores. My a8x only gets a 2700?

Great app

between all the other less useful benchmarks, this one is really standing out...

Is it accurate?? Im getting a 4617 with quad selected on my lg g2. I try running linpack on my ipad air 2 (a8x) and only get 2650. Maybe i need to compare with the other versions of linpack. The a8x is much faster than the snapdragon 801 especially per thread.

Thanks

between all the other less useful benchmarks, this one is really standing out...

Unlike with Linpak, the results do not wildly vary between separate runs. The MFlops result are also directly related to your CPU speed (with half CPU frequency, my score was close to half as well) so you can easily tell if that particular subsystem has improved.

The app delivers consistent results and doesn't seem to be terribly affected by kernel revisions or optimizations. The app works hard to do that, too - it keeps the screen awake so you don't have to constantly give input to keep it running at full. Very useful metric to compare various tablets and phones.

I don't know how this app works, or how does Linpack work, but if they can be compared, and linpack bencmarks the dalvikVM with a similar algorythm, this application shows, that running everything in a virtual machine, is a very bad idea. It didn't solve any compatibility issue, or the manufacturers still do not update their phones. Our phones are getting more and more powerful, and we can use about 10% of the maximum performance. Use native C++, and forget dalvik. Application is stable, the test show constant results. Also a good feedback for overclock performance. Perhaps, it needs an 8 thread mode soon. The asus padfone 2 is here... Also, if there is a background activity, it stops benchmarking. (like receiving a call, sms, skype PM)

Samsung galaxy s2 x. 1 thread, 505.0 mflops what ever that means,

Xperia u ICS

It works. Stock Nexus 4: 2853, 1499, 764 using 4, 2 and 1 threads.

I am so excited for this app as it is an accurate test of the FLOP performance of a CPU unlike Linpack for Android that is so often used. It is simple and consistent, and provides great insight into the performance of your device. I am looking forward to a renderscript test! ;)

Works fine on Motorola RAZR i.

Provides extremely consistent scores and runs fairly quickly, allowing me to easily benchmark different ROMs and power profiles.

Application. The results are consistent and reproducible and this is very important when comparing devices. 2337.0 on the Galaxy S III on JellyBean.

Excellent!

Awesome

Works fine upon Nexus 7 tablet Results: 1 thread - 316.0 MFlops 2 threads - 880.0 MFlops 4 threads - 1454.0 MFlops Would like the option to have other thread counts selectable, 5 on a quad core would be interesting in testing other aspects and in that feel the selection of 1,2,4 whilst useful a larger selection would of made for a more flexable test.

works It works. Stock Nexus 4: 2853, 1499, 764 using 4, 2 and 1 threads.

259.0 x1 thread Xperia u ICS

It's native I don't know how this app works, or how does Linpack work, but if they can be compared, and linpack bencmarks the dalvikVM with a similar algorythm, this application shows, that running everything in a virtual machine, is a very bad idea. It didn't solve any compatibility issue, or the manufacturers still do not update their phones. Our phones are getting more and more powerful, and we can use about 10% of the maximum performance. Use native C++, and forget dalvik. Application is stable, the test show constant results. Also a good feedback for overclock performance. Perhaps, it needs an 8 thread mode soon. The asus padfone 2 is here... Also, if there is a background activity, it stops benchmarking. (like receiving a call, sms, skype PM)

works well i guess Samsung galaxy s2 x. 1 thread, 505.0 mflops what ever that means,

Awesome

Excellent Application. The results are consistent and reproducible and this is very important when comparing devices. 2337.0 on the Galaxy S III on JellyBean.

Consistent results, convenient to use. The app delivers consistent results and doesn't seem to be terribly affected by kernel revisions or optimizations. The app works hard to do that, too - it keeps the screen awake so you don't have to constantly give input to keep it running at full. Very useful metric to compare various tablets and phones.

Simple and works as outlined Works fine upon Nexus 7 tablet Results: 1 thread - 316.0 MFlops 2 threads - 880.0 MFlops 4 threads - 1454.0 MFlops Would like the option to have other thread counts selectable, 5 on a quad core would be interesting in testing other aspects and in that feel the selection of 1,2,4 whilst useful a larger selection would of made for a more flexable test.

Extremely consistent Provides extremely consistent scores and runs fairly quickly, allowing me to easily benchmark different ROMs and power profiles.

meaningful benchmark between all the other less useful benchmarks, this one is really standing out...