Interesting developments this week as HP, Intel and Yahoo (annointed the title "Yahooptel" by Allistair Croll--see GigaOm post linked later) announce a collaboration to build a cloud targeted at academic and research organizations. Defined, perhaps, as a grid service cloud, the offering is well documented by the aforementioned GigaOm post, William Vambenepe, and the world, so I won't go into too much detail here.
However, I do think there is an interesting little trend here, noted by Stacy in her GigaOm post. Google and IBM, along with Yahooptel are pushing into academic circles for--at least in part--a very interesting reason: to allow computer science and IT Management students to gain experience building applications in so-called Internet scale architectures. The battle for young minds has begun, and the big dogs have a huge advantage here.
Now, there is an interesting counter force at work that will temper any chance of "total domination" by the large established players; the most successful cloud computing is based, at least in part, on open source. So the architectures that tomorrow's SaaS superstars deploy may be heavily influenced by a large vendor, but the frameworks, languages, interfaces, protocols and perhaps even services will be freely available from an open source project and/or vendor.
This is the beauty of open source in the cloud, at least for software. It firmly removes any great advantage to being an established player in the preceding market. The code it takes to build the next generation architecture becomes free and accessible, meaning anyone can get their hands on it.
Hardware, on the other hand, is still expensive to assemble in large configurations. I think this is the real driver for these academic clouds--not necessarily to lock in proprietary software approaches for cloud computing, but certainly to sell a hardware platform. Isn't that half the battle for a high-scale system developer's soul?
