OK … did an analysis using Topofusion's climbing analysis tool
http://www.topofusion.com/climb.php
The result is this
The data is smoothed using Topofusion's alogrithm that uses interpolated smoothing of GPS altitude, distance subdivision and minimum gain. Other methods are used too, such as DEM (Digital Elevation Model) which makes use of 3-d models of topographic maps. The methods all give total ascending of between 7600 and 8100 feet
The climb data is:
TopoFusion: 8136 ft
Straight GPS: 8140 ft
Straight DEM: 7605 ft
TopoUSA: 7638 ft
So it appears that 3500-2500 ft or so is noise …
GPS distance and time summaries:
Moving time 6:47:18
Distance: 29.1 mi
Average Speed: 4.3 mph
Max Speed: 36 mph
Uphill Time: 2:37
Downhill Time: 2:04
Flat time : 2:06
Uphill dist: 11.02 mi
Downhill dist: 10.94 mi
Flat Dist: 7.00 mi
OK … not exactly Alpe d'Huez, BUT, for what its worth, if you take the distance ridden uphill and divide by the elevation change over the distance ridden uphill, by the least generous measure (7000 ft in 11 mi), you get a grade of just over 12% … that would be like 1.5 Alpe d'Huezs at its max (reported) grade. So of, enuff with the Alpe d'Huez comparisons.