Datacenters In The Ocean?
So, I have to admit, in re-reading the last few posts of mine, they seem a bit dry (although the news has been big!). I've been considering how to overcome my lack of access to the inside scoop compared to one of my favorite bloggers. Since Jonathan's blog is my favorite blog of all time (I'm not kidding and I swear I'm not sucking up), I recently realized I've been part of some cool conversations.
I have lived (note: I'm not saying "spent") the last 9 years as a Sun Sales Representative. It has been a blast. After all, we have IP at all levels from a systems perspective (remember Scott McNealy years ago reminding the marketplace we are Sun MicroSYSTEMS?). I've truly enjoyed representing Sun at our customer sites advocating why we're so relevant and cool. I've worked with partners, small startups, and established enterprises. I've been in a lot of meetings (we're reporting our customer face time, since we are laser focused on how we can help customers.....so I hesitate to say I've been in too many meetings :-) In any case, I've seen and heard alot. It dawned on me, there's one topic that always seemed to catch my attention....."The Wall".
One of the last accounts I had a daily focus on was growing by leaps and bounds (and still is). They couldn't buy systems fast enough. I've been involved in some recent customer meetings for other accounts, and I've heard this is not a unique sentiment. How can customers build quick enough? Should they put systems literally anywhere they have space? I was in a customer meeting recently where they said they even considered putting these datacenter modules in international waters! (I was impressed to learn they thought through many issues including power and warming of the oceans!). customers are worried about facing hitting "The Wall" of capacity capability. At some point, servers and datacenters aren't big enough, then what? The bottom line is that as the web buildout continues, Sun's competitors must deliver some value through their products (I forgot to mention we spend ~$2B on R&D so you don't have to worry about how to leverage network infrastructure to keep cost and complexity down!) as these issues can't be solved without an engineering focus...a distribution model just won't cut it when throughput, power, support, and performance matter as you focus on your competitive edge. I'll write in more detail in the coming week or two about some specific technologies, like Sun Modular Data Center, that web companies are considering in meeting this demand. The buildout has really just begun......