Archive for September, 2007|Monthly archive page

GPU Labeling is a Necessity, not a Luxury

The new ESRI Streets and Data 2007 Update now includes Streetmap Pro. It’s free, but you need to request it.

I’m trying to use StreetMap Pro data with a tracking application. The labeling is a bottleneck, to see what I’m talking about do this:

Open up the StreetmapPro.mxd that ships with ESRI Streets and Data 2007 Update. Then open up the task manager on CPU history. Pan around the map at 1:500K in an area with lots of labels. Try it with labels turned off, then with labels turned on. Notice how much extra cpu is burned when the labels are turned on.

There must be ways ESRI could speed this up. It should be possible to leverage the GPU to do this. Figuring out how to place labels so they don’t overlap is tricky, but seems parallelizable.

I wonder if Google Earth will be using Peakstream to do this. This webinar looks exciting, until you realize that shortly after it was released, Google bought Peakstream and locked them in a closet. Now thats evil.


Wiki City Rome

Web Urbanist has an interesting post on Wiki City Rome. I find the data a bit too live, would be more interesting to include heatmaps that change slowly, more like a lava lamp.

Spinning Globes – killer apps for Multicore?

AMD has released a new quad-core processor, dubbed Barcelona, but as with Intels offering, there seems to be a consensus that until software starts leveraging multicore, demand will not take off as hoped.

Seems like geoprocessing would be a natural fit for running on multiple cores. From what I’ve seen though, writing software for multicore is hard. Google bought Peakstream a couple of months ago, a company that makes software tools for multicore development. Here’s a good description. Unlike their acquisition of Sketchup, though, Peakstream has been pulled out of the public eye. To me this means Google thinks Peakstream gives them a competitive advantage. That’s too bad, I really wish there was some sort of free download of PeakStream that would allow me to start playing with multicore, writing tools that extend GE.
Maybe Google could even enhance KML with tags indicating how multiprocessors should be exploited. Think of a dynamic 3D weather model described in KML so that load is spread across multiple cores as it runs.

Firing up Google Earth Beta on my dual core doesn’t very evenly tax each core, but I wonder if this will change once Peakstream gets digested. Once that happens, could Google Earth spur demand for multicore?

In looking at AGX, it seems like the custom task framework could be extended to recognize & exploit multicore. ESRI has done a nice job at hiding the threading complexities, seems like the custom tasks framework could be extended so that I can tell a custom task how to handle and choose between multiple cores.