How accurate is Quadrant?

[Demo24]

I decided to test my D1 just to see where it stacks up these days with the overclock and different rom. But I keep getting wildly different scores. Sometimes in the mid 800's to even a 1130 at one point, but most tend to be around 1050 to 1100. I understand a few points, but it seems to vary too much to be random variance. Anyone else have such differences or does your phone tend to score roughly the same on each pass?

 

[Bateluer]

I've had a few flukes putting my D1's score into the mid 2K range. Never reproducible though. Depends on the ROM and the OC, but I wouldn't put too much stock on Quadrant. Its been shown to be relatively unreliable.

 

[Oyeve]

I wouldnt put much faith in quadrant. My Vibrant on 2.2 stock scores rather low.

 

[Dulanic]

Quadrant is horrible horrible horrible. With a few kernel tweaks that have 0 real world performance improvement, you can jump a score 500+.

 

[ChronoReverse]

It also tests too many things besides the CPU so that if your phone has one particular thing it's good at, it'll boost the score like crazy.

 

[shabby]

Quadrant is horrible horrible horrible. With a few kernel tweaks that have 0 real world performance improvement, you can jump a score 500+.

I got my captivate to hit 2400 on quadrant, yet it still has minor hiccups/freezes like the stock one that only manages 1200. Once you start tweaking the phone it seems like quadrant scores become meaningless.

 

[dguy6789]

Quadrant results are extremely easy to cheat and aren't a good indicator of phone performance at all. It's about as relevant to phones today as 3DMark2001 is to PCs today. It is used a lot because it is full of features but people ignore the fact that the end results are meaningless and at the end of the day that is what matters.

 

[mnewsham]

2800 On DroidX

But honestly it doesn't really mean crap.