EC2

I've been thinking a lot about Amazon lately. The Elastic Compute Cloud is changing how consultants provide services to small business and deploy custom solutions. A couple years ago, I would have sold a VMWare farm to a small business office, offering excellent resilience and capabilities in a small footprint. Today - I would never sell that solution. The risks and infrastructure requirements for even a half-rack solution are far too expensive compared to solutions deployed in Amazon.

My home lab is a good example. Last winter I scoured eBay and invested in some servers coming off a lease. I have enough compute power and storage to run 10 virtual machines and store 4000 DVDs, and it cost less than $5k. (I chose a hardware solution for my lab to better understand virtualization.)

But for a busines customer, I would have built the environment in Amazon for a couple hundred a month. It would be 2 years before I hit my up-front costs, and I would never have to worry about power, air conditioning, insurance, security, wiring, or backups.

As a consultant, it is a revolution! I don't have to invest my capital to build a solution, and then wait the Net90 to be reimbursed. I setup accounts in my customer's name, with their credit cards, and I just bill for services. I have less exposure to risk, and I can sell a more robust solution.

Translating this to the enterprise is still a tough sell, but for independant IT folks, Amazon makes too much sense not to deploy.

My take on HP's WebOS decision

HP's recent decision to shutter their WebOS and tablet division strikes me as extremely short-sighted. Spinning-off the PC busines strikes me a smart move because PCs are on a path to extinction, but HP now has no in-house solutions for Tablets or Mobile phones. In a future dominated by ubiquitous computing and connectivity, having in-house knowledge, platforms, and intellectual property are necessities. HP hasn't moved fast enough to deploy consumer-oriented Internet services (aka Cloud) which might help it maintain long-term market dominance, so having client platforms for cloud services seemed like a real opportunity for long-term advantage. And HP needs to purchase a lot of companies if they hope to compete with software and service giants Oracle and IBM.

Short-term, HP will probably recoup their costs by selling Palm's patents, but it's hard for me to understand how they'll deploy that capital in their existing businesses to better effect than the investment in mobility.


P.S. See Al Lewis' column in WSJ. "H-P's One-Year Plan" highlights all the stuff I forgot. What is going on at HP?