Facebook’s new data center in Prineville, Oregon, is full of Fusion-io technology, and the one planned for Rutherford County, North Carolina will be as well. Fusion-io CEO David Flynn told me in an interview some time ago that servers running Fusion-io technology constitute about 80 percent of the hardware footprint inside the Facebook building. If you ever wondered what kind of secret sauce Facebook might have in its data center infrastructure, make no mistake: Fusion-io’s flash memory-based technology is it.

The close links between Facebook and Fusion-io have a lot to do with a mutual relationship. Venture capitalist Marc Andreesen sits on Facebook’s board and is an investor in Fusion-io. Goldman Sachs and Morgan Stanley are leading the offering with JP Morgan and Credit Suisse. The filing doesn’t say how much the company expects to raise in the offering.

Along with Facebook, Fusion-io’s hardware partners Hewlett-Packard, and IBM also account for more than 10 percent of revenue.

This filing is a surprise, because as recently as December, Flynn told me the company was in no hurry to go public. It had just raised a $45 million C-round from Meritech Capital Partners, with Accel Partners, Andreessen Horowitz, New Enterprise Associates, and Triangle Peak Partners also invested. Plus it has strategic investments from Samsung, the South Korean electronics and semiconductor giant–which also happens to be the world’s biggest supplier of the flash memory chips key to Fusion-io’s technology–and Dell Ventures.

But from the looks of the numbers, it appears the company needs to raise the kind of cash that an IPO can provide. The company finished 2010 with $36 million in revenue and a $31.5 million net loss despite a gross margin of 59 percent in the latest quarter. Operating expenses were nearly $52 million, led by $24 million in sales and marketing expenses. The balance sheet shows $3.5 million in cash and cash equivalents plus another $39 million in working capital. That’s not enough to fund operating expenses assuming they remain at the same level, which they won’t, given how fast the company is ramping up.

And ramping up it is. Fusion-io just last week announced a relationship with its fourth major server manufacturer, the white-box manufacturer Supermicro, and on the same day, IBM doubled the number of servers that include Fusion-io as an option. Before that, it scored a rare public disclosure from Credit Suisse that its technology was being used to speed up trading, and there are certainly more financial companies buying the technology and simply not talking about it.

Sounds like another one of those poorly conceived “20-percent time” projects, doesn’t it? Historically, corporate venture capital portfolios have a very mixed record. There are the adepts of the discipline–Intel Capital (INTC) and Johnson & Johnson Development Corp. (JNJ), for example–that have done well for their parent companies. And then there are the maladroits like Dell Ventures (DELL) and Applied Materials Ventures (AMAT) whose crowning achievement was their ignominious retreat from corporate VC in 2005. A risky and often thankless proposition, running a corporate VC shop. “… Venture investing is not the best use of a corporation’s capital,” says venture capitalist Fred Wilson. “It is inevitable that it will produce sub-par returns at best and significant losses at worst.”

So why bother? Well, if you’re a company that takes the long-term view, as Google (GOOG) does, you’re likely more concerned with long-term strategic goals than short-term financial ones. So why not take some of the vast wealth you’ve accumulated in pursuit of those long-term gains, invest it in some start-up’s new ideas and innovations and incubate them–short term? The start-up company might be the next you, right? And if it is, you acquire it (presumably, your VC investment deals include first-acquisition rights). And you do so at a price lower than the one it would fetch after taking funding from traditional VCs. Then you bring the Google hive-mind and infrastructure to bear on the start-up’s innovations and bring them to market or use them to enhance your own product and services.

If the start-up turns out not to be the next you, you don’t follow up on your initial investment–saving yourself the grief of a dud acquisition.

Essentially, you transform the emerging technology market into one vast Google Labs from which you have pick of the litter. How’s that for a 20-percent time idea?

[Image Credit: VC Wear]

Sounds like another one of those poorly conceived “20-percent time” projects, doesn’t it? Historically, corporate venture capital portfolios have a very mixed record. There are the adepts of the discipline–Intel Capital (INTC) and Johnson & Johnson Development Corp. (JNJ), for example–that have done well for their parent companies. And then there are the maladroits like Dell Ventures (DELL) and Applied Materials Ventures (AMAT) whose crowning achievement was their ignominious retreat from corporate VC in 2005. A risky and often thankless proposition, running a corporate VC shop. “… Venture investing is not the best use of a corporation’s capital,” says venture capitalist Fred Wilson. “It is inevitable that it will produce sub-par returns at best and significant losses at worst.”

So why bother? Well, if you’re a company that takes the long-term view, as Google (GOOG) does, you’re likely more concerned with long-term strategic goals than short-term financial ones. So why not take some of the vast wealth you’ve accumulated in pursuit of those long-term gains, invest it in some start-up’s new ideas and innovations and incubate them–short term? The start-up company might be the next you, right? And if it is, you acquire it (presumably, your VC investment deals include first-acquisition rights). And you do so at a price lower than the one it would fetch after taking funding from traditional VCs. Then you bring the Google hive-mind and infrastructure to bear on the start-up’s innovations and bring them to market or use them to enhance your own product and services.

If the start-up turns out not to be the next you, you don’t follow up on your initial investment–saving yourself the grief of a dud acquisition.

Essentially, you transform the emerging technology market into one vast Google Labs from which you have pick of the litter. How’s that for a 20-percent time idea?

[Image Credit: VC Wear]