Den amerikanska bloggen Android Police har haft chansen att testa en Nexus One med Android 2.2/FroYo och bland annat kört benchmarktestet Linpack for Android.

Med Android 1.6/Donut så presterar Xperia X10 4,1 MFLOPS och med Android 2.1/Eclair så presterar Nexus One cirka 7 MFLOPS och nu till det spännande, med Android 2.2/FroYo presterar Nexus One hela 37,5 MFLOPS. En ökning som enligt våra tidigare resultat motsvarar en förbättring om 528%, inga dåliga siffror med andra ord.

Varför prestandaskillnaden är så stor beror på att Google med Android 2.2/FroYo ryktas ha implementerat en JIT-kompilator, något som i situationer som ovan bevisligen innebär en kraftig prestandaförbättring. Däremot skall man inte förvänta sig en så stor förbättring över hela brädet, som beskrivet i utdraget från Wiki-artikeln nedan så konverterar JIT-kompilatorn bytekod till maskinkod, något som innebär att långt från alla applikationer/operativsystemet i sig kommer att köras snabbare. Android Police säger dock att telefonen med Android 2.2/FroYo känns och upplevs som bra mycket rappare än med Android 2.1/Eclair.

In computing, just-in-time compilation (JIT), also known as dynamic translation, is a technique for improving the runtime performance of a computer program. JIT builds upon two earlier ideas in run-time environments: bytecode compilation and dynamic compilation. It converts code at runtimeprior to executing it natively, for example bytecode into native machine code. The performance improvement over interpreters originates from caching the results of translating blocks of code, and not simply reevaluating each line or operand each time it is met (see Interpreted language). It also has advantages over statically compiling the code at development time, as it can recompile the code if this is found to be advantageous, and may be able to enforce security guarantees. Thus JIT can combine some of the advantages of interpretation and static (ahead-of-time) compilation.

–Wikipedia.org Just In Time Compilation

Linpack är ett program specifikt framtaget för att utvärdera prestandan i Dalvik VM (virtuella maskinen) och nedan står beskrivet hur det hela fungerar.

The LINPACK Benchmarks are a measure of a system’s floating point computing power. Introduced by Jack Dongarra, they measure how fast a computer solves a dense N by N system of linear equations Ax = b, which is a common task in engineering. The solution is obtained by Gaussian elimination with partial pivoting, with 2/3*N³ + 2*N² floating point operations. The result is reported in Millions of FLoating-point Operations Per Second (MFLOP/s, sometimes simply called FLOPS).

This test is more a reflection of the state of the Android Dalvik Virtual Machine than of the floating point performance of the underlyingprocessor. Software written for an Android device is written using Java code that the Dalvik VM interprets at run time.

What is the purpose of this app?
This is a simple benchmark test to show performance relative to other phones for a standard calculation.  Linpack has been used for years on all types of computers, with a version used to rate the TOP500 computers in the world.

What speed is better?
A higher number is better.

Does having faster speed improve the android phones or what?
Yes, it should.  The Dalvik VM has a huge impact on the Linpack number.  A better number on the same device would indicate that a new version update has improved performance.  Or it could show that something has gone terribly wrong if the number goes down.

—Greenecomputing

Genom Android Police