Over the last couple days, I've had the pleasure of speaking to the customers of an ISV that offers applications which perform deep analyses of massive bodies of data harvested from scientific lab equipment (like mass spectroscopy and liquid chromotography machines.)
While I don't know much about what all that means (despite watching copious CSI Miami reruns), I do know that many of the people I spoke with at the conference:
1. Didn't know that Microsoft has a high performance computing offering
2. Didn't realize that it was a low-cost solution designed to run on commodity hardware
3. Didn't realize that it integrates nicely with the Excel Calculation Service
Perhaps none of that should have surprised me, but I was a bit taken aback by my underestimate of the prevalence of Excel as an analysis tool within the scientific ranks. Well, I don't know if I'll ever be anything more than a wanna-be biogeek, but I can tell you that it doesn't take rocket scientist to see the opportunity here for enterprising ISVs. The prospect of running complex spreadsheet calculations (or other resource intensive computations) in a commodity grid makes scientific types really eager to give it a try - and with a little additional architecture, ISVs can make these solutions inherently CPU load-balanced and inherently route around faults in the grid fabric.
If you're a scientific ISV that's at all interested in exploring these kinds of solutions with us, let us know (because I can assure you we're interested in helping you do so.) And as an added incentive, if you can bring a scientific app to market on our grid technologies, I'll see if my boss can be convinced to send you a Zune with "Blinded by Science" pre-loaded (no guarantees though...)