There was a time at Sun Microsystems, not so long ago, when news of a $700 million investment by leveraged-buyout masters KKR would have triggered talk of a sale, break-up or a brutal change of leadership. In fact, for years, many analysts have been expecting just that, saying that sooner or later investors would give up on the company’s expensive passion for R&D and its self-appointed mission as Silicon Valley’s most visionary company.

But not now. Since July 2006, Sun’s share price has jumped by more than 40%, helped by an unexpectedly strong and early swing to profits after years of losses. In its most recent (second) quarter, Sun reported sales up 7% to $3.56 billion, with profits of $126 million. For the year leading up to June 2007, Sun is expected to turn in net profits above $1 billion, with sales topping $14.85 billion.

These latest quarterly profits don’t seem stellar, but they need to be seen in context. For five straight years, from 2001 to 2006, Sun turned in heavy annual losses totalling $5.3 billion; sales fell for the first four of those years (see chart). Without an extraordinary and unexpected $2 billion payment from Microsoft in 2004, to settle a long-running series of anti-trust disputes, many observers think Sun’s longstanding CEO Scott McNealy (now chairman) would not have survived, and that Sun would have surrendered its independence.

But although credit for Sun’s turnaround clearly owes much to McNealy’s persistence and his insistence on maintaining technological leadership, most of the praise will go to his pony-tailed heir Jonathan Schwartz, who took over the top job in April 2006 – just as the company’s fortunes were turning round. “Schwartz and his team have demonstrated remarkable vision and strong discipline,” said KKR in making its minority investment.

Peter Ryan, vice president for EMEA at Sun, says its recovery can be largely attributed to Sun’s long-standing investment in R&D, which at between 15% and 19% of sales is way above industry averages. Often against prevailing trends and external advice, Sun has continued to fund development in processors (Sparc), operating systems (Solaris), thin client devices (Sun Ray) and software development (Java).

With so much high-profile activity, it is sometimes forgotten that Sun’s real business is in hardware, which accounts for 63% of revenues (including storage). If there is one area where its investment is paying off, it is in the latest range of ‘Niagara’ UltraSparc multi-core processor powered servers. “This is the fastest ever $0 to $100 million business in the company’s history,” says Ryan.

Although Sun makes much of the fact that its Solaris operating system (along with Java) has been made available as an open source product, it is not that which is driving server sales. The Niagara processors have been designed with enterprise workloads in mind, and the servers are simply more powerful, give out less heat, and use less power than alternatives from rivals. In addition, after years of resisting, Sun has at last acknowledged that some customers want low-cost Intel- and AMD-based servers.

With these two business lines booming, Sun’s server revenue is currently growing at over 25% a year. With environmental issues set to influence IT buying strategies in the coming years, some analysts think its underperforming Sun Ray thin-client business will soon take off, further boosting server sales.

But what of its two big acquisitions in 2005 and 2006 – SeeBeyond, the integration software specialist, and StorageTek, the storage company focused on tape? These purchases helped boost revenues in 2006, but it is still unclear whether Sun has been able to dramatically strengthen its position in either of these two markets. But, says Ryan, each had large customer bases in areas where Sun was less strong; at the very least, they have opened up new client opportunities.

Sun’s renaissance, like that of Apple, is being seen in Silicon Valley as a vindication of deep R&D and a commitment to innovation. With so many customers having to deal with really big IT problems, says Ryan, “it’s important somebody understands the whole stack [from processors upwards]. In the [recent] past, that’s been delegated to systems integrators. But they are focused on deploying, not innovating.”