Linpack
Nexus One with Froyo takes the Lead at 40.099 MFLOPS
Mar 15th
We’ve had two different Nexus One phones with four good Top 10 Linpack runs. The best being 40.099 MFLOPS! Looks like some Droids need to do some catching up. I’m fairly certain that these are not bogus runs, since I’ve added some anti-cheat mechanisms to the most recent release. Still some funny results do occur when you overclock.
So is it Google teasing us??
As cyanogen said via twitter “Linpack #4 is an N1 running Froyo. Cheaters”
And of course, TheDudesAndroid said “***shakes fist at userid****”
XDAndroid
Mar 12th
TMONEWS has a posting about XDAndroid and how it is much more stable on a HTC Touch Pro 2. As they state, “the Android installation experiences no difficulty in making calls, sending SMS, downloading apps, or browsing the web. In fact, pretty much everything works besides GPS, Bluetooth, and the camera. ”
XDAndroid is a project to port Android to Windows Mobile devices. This is being developed by the guys on the forums of xda-developers.
What I found interesting was that there were a couple Linpack postings for XDAndroid loaded devices. So, I added XDANDROID to the Device list. It will be interesting to see how well it performs on Windows Mobile devices.
Removed Jw52637 50.592 Run from TOP10
Mar 11th
Ok, sorry Jw52637. I decided to delist the 50.592 run from the Top10 list. I’ve received and seen lots of comments from people that the run was faked, that you hacked into the list or I just made it up. From what I can tell it was a legitimate Linpack run since I have a check for fake uploads. I also doubt the site was hacked to add it since it also had the right check value. About the only thing I can think, that it didn’t run correctly and thus had an incorrect MFLOP number. I say this since his runs the day before topped out at 9.269
So in the next version, I’m going to check for an invalid run so that it can’t get uploaded.
Friendly little Linpack War
Mar 6th
Taking a few days off from here because of my “real” job was hitting the fan. I see that we have a Linpack War going on. If you look at the Top 10 list and the Nexus One device list you’ll see lots of movement. AndroidSpin has done a nice job of covering all the bases on the reasons behind the friendly competition.
“Originally this was just a “friendly” little bet between WootRoot and TheDudesAndroid that a 2.0GHz kernel would or wouldn’t work on the Nexus One. Somehow WootRoot convinced Kmobs to build the kernel (I have a feeling Chris Soyars also played a part in that) and WootRoot bet TheDudesAndroid $50 that he could boot his Nexus, send a tweet and play a video before his phone melted.
WootRoot tried. It didn’t work. But it also didn’t melt. Kmobs got curious and toyed around with frequencies and voltages and they went from 1.113GHz (which was supposed to be the stable max) to 1.267GHz. Of course, it wasn’t until Acsteffy87 decided to run the kernel on his Nexus as well, that TheDudesAndroid found his “courage” and booted it on his too. (I think I’m going to be in trouble for that, lol.) At first, only The Dude’s Nexus would overclock to 1.267GHz and for that reason, it was dubbed the “JesusPhone” or JP1.”
Fixing things with Linpack App and Reporting
Feb 21st
I’m slowly getting some things fixed with the Linpack for Android app. I’m adding information about MHz used and hopefully some reporting directly into the app on where it scores on the master list.
I’ve also fixed a big error on my part dealing with MFLOPS. For some reason, I had it listed here and in the app as Mflops/s. Oops, that would mean Million of Floating Point Operations per second per second. It would be cool to measure acceleration, but that is not what it reports.