How accurate IS Quadrant?

[Dknelson728]

Hey guys, I'm just curious about the accuracy of Quadrant for figuring out the benchmark speed of one's Droid.

I stopped by a Verizon kiosk today, and went ahead and installed / ran Quadrant, just because I was curious about the result from the Droid 2.

It did it's test, and came back with a result of 1.3

Seems a bit high to me, maybe someone could shed some light?

Cheers!

 

[Dave12308]

???????

You're going to have to explain yourself a bit better, as Quadrant scores have nothing to do with GHz.

It should be a whole-number score, something like 1300 for a Droid 2, IIRC.

So, if the score was 1300 and that's what you meant by 1.3GHz; then it's just about right.

 

[Dknelson728]

My mistake, I have two thoughts running at once here. That GHz shouldn't have been there. ^^
I was just under the impression that the benchmark number for the D2 was generally lower than that.

 

[OSX2000]

I don't think Quadrant is accurate at all. Check out the two scores in my sig...those were both done with the same ROM and kernel, UD7 with Chevy's 1.1GHz. With Stagefright on, the H.264 and AAC decoding tests get skipped. This results in a higher score, because the tests were instantanious. With Stagefright off, the tests are done correctly, so the score is cut in half.

 

[Dave12308]

My mistake, I have two thoughts running at once here. That GHz shouldn't have been there. ^^
I was just under the impression that the benchmark number for the D2 was generally lower than that.

I'm not too familiar with the Droid 2; but it is an Android 2.2 device; and it has a 1GHz processor. I will admit that I am speculating a bit here, but my Droid 1 is also running 2.2 - and i've overclocked my processor to 1GHz. And I get ~1400 - so if anything, 1300 is a bit on the low side considering improvements have been made to the Droid 2's chipset.

 

[Dave12308]

I don't think Quadrant is accurate at all. Check out the two scores in my sig...those were both done with the same ROM and kernel, UD7 with Chevy's 1.1GHz. With Stagefright on, the H.264 and AAC decoding tests get skipped. This results in a higher score, because the tests were instantanious. With Stagefright off, the tests are done correctly, so the score is cut in half.

The true problem isn't with the accuracy of the benchmark; it's because the dev of the app needs to update it to be stagefright-aware.

Synthetic benchmarks are seldom of any real use anyways, aside from "bragging rights". It is rare that a synthetic benchmark even comes close to showing real-world performance.

 

[harrellj]

Quadrant is still able to be faked out though, there's screenshots of a CM6 running Droid with a score in the 3000s.

 

[rwking]

Quadrant is still able to be faked out though, there's screenshots of a CM6 running Droid with a score in the 3000s.

That's insane. I was also curious about this, and came across this thread. My Droid is coming back with results of 900 - 1250. Why the disparity? That's what makes me dubious.

 

[Darkseider]

There are a lot of ways to get higher results in Quadrant than what you should be getting. A friend with a Captivate using the speed fix and OC'd to 1.2 Ghz is scoring 2300's. Granted the Galaxy S is a fast phone... not that fast. Did I mention that it is still 2.1?