VFSMark Scores

[Andyvan]

I ran VFSMark on a TG50 at Circuit City the other day. While the overall mark was much better than my N610, some of the individual scores were much worse.

Sorry, I don't have the TG50 scores with me.

Just for reference, my N610 has a VFSMark of 71. I believe the TG50 was either 140 or 170.

Any ideas why the TG50 is so much faster on some aspects (reading), and so much slower on others (writing?)?

-- Andyvan

[YTTAN]

Different chip got different speed. Maybe the read and write chip is diff within a set.

[s_n_m]

VFSmark is an os 4.0 app, therfore runs in emulation.

If the app were os 5 native (andonly ran in OS 5) you would not only get an accurate reading, but you would be suprised just how fast a 200mhz xscale chip really is.

[kkerruish]

Just the CPU speed alone will account for why it's so much faster...200mhz vs 33mhz...

But s_n_m's post regarding the application's code explains why the results were so weird...

Ken

[Andyvan]

You'd expect that memory stick speed compared to emulation speed would be a large ratio, hence any emulation overheard would be mostly unnoticable.

Maybe the test code needs to be updated.

-- Andyvan