Frank is back.

Frank Quattrone, Silicon Valley’s most powerful investment banker in the 1980s and 1990s, picked a moment of maximum market turmoil to announce that he’s back in business. Yet despite predictions he’d start a private-equity firm, Quattrone instead is returning to his first love, straight-on investment-banking services to high-technology firms. His new outfit, Qatalyst Group, will start as a six-partner boutique in the mold of Greenhill & Co. (GHL), Evercore Partners and Moelis & Co., all firms started by former bankers at high-profile firms.

Another prediction that didn’t pan out: Quattrone’s new firm won’t include his former partners, George Boutros and Bill Brady, who together with Quattrone dominated the tech banking world for more than a decade as the team traveled from Morgan Stanley to Deutsche Bank to Credit Suisse, where Boutros and Brady remain. His five founding partners at Qatalyst are a group of 20- and 30-somethings, each of whom worked with Quattrone at Credit Suisse, though none was there immediately before joining Quattrone. The five are Jonathan Turner, 34, a former Internet banker and most recently a biz-dev executive at the online marketing company QuinStreet; Adrian Dollard, 38, the firm’s general counsel; Neil Chalasani, 29, who did a stint at Evercore; Brain Slingerland, 30, who decamped to Goldman Sachs after Credit Suisse; and Brian Cayne, 26, who came from Vista Equity Partners.

For a while, it looked like Quattrone’s name would be linked with the likes of Dennis Kozlowski and Jeffrey Skilling, both of whom are doing time in jail for crimes committed during the market mania that surrounded the dot-com craze. Yet Quattrone’s conviction on obstruction of justice was overturned and he was fully exonerated in 2006. He says he’d been thinking about starting a private-equity firm but decided instead to focus on what he knows best. “I’m more of a growth guy and a strategy guy,” he said, during a Tuesday-morning interview from his firm’s temporary offices in San Francisco.

For all the negative press Quattrone got during his trials, his support base in Silicon Valley remained remarkably strong. It showed in the big hitters he lined up for his firm’s inaugural news announcement. Google (GOOG) CEO Eric Schmidt, Intuit (INTU) Chairman and Valley consigliere Bill Campbell, Facebook investor and venture capitalist Jim Breyer, and Facebook CFO and former Yahoo (YHOO) treasurer Gideon Yu each lent their names to enthusiastic testimonials.

Quattrone says the new firm has no clients yet as it awaits approval of its broker-dealer registration, a process that could take up to six months. In the meantime, Qatalyst will operate as a division of JMP Securities (JMP), much the same way former UBS banker Ken Moelis operated initially as part of Mercanti Securities. Indeed, Moelis is more than a role model for Quattrone. He’s an example the kind of business Qatalyst hopes to win. Moelis currently is advising Yahoo on its defense of a Microsoft (MSFT) takeover bid, precisely the kind of assignment Quattrone wants to be in the position to take on. Qatalyst also will raise a fund for investing alongside its clients, though Quattrone says that initially the money will come from himself and his partners.

Quattone says that after some “soul searching” he realized that he doesn’t miss the empire-building and “liasing” with New York, Germany and Switzerland that went along with running outposts of major banks during the years he and his team backed iconic companies like Cisco (CSCO), Netscape and Amazon.com (AMZN). What he misses, he says, is giving “good, old-fashioned, honest advice.”

While Quattrone has been taking time to reflect, of course, his former minions have sprinkled themselves throughout Wall Street. Watching him and his new young recruits compete against them will provide some good, old-fashioned fun in Silicon Valley.

I saw something more than a little scary last week at Google (GOOG): A calm, confident, friendly management team that seemed more comfortable in its own skin than I’ve ever seen them.

For the first time, Google invited a handful of journalists to its annual Zeitgeist conference for advertising partners. Afterwards, Larry Page, Sergey Brin and Eric Schmidt held an on-the-record chat with us over sandwiches. They talked about their interest in wireless spectrum. They fretted that their biggest challenge (still) is managing their growth. They even showed their sense of humor. When I noted the unusual stability of the top executives beneath the ruling troika (as a prelude to asking if that stability would continue), Page quipped, “Emotionally or physically?” More seriously, he predicted continued stability and suggested a leadership training program is helping. (The word is that many of Google’s “economic volunteers,” a term that’s actually used within the company, will be retiring soon and that turmoil will hurt.)

The arrogance remains, of course. Schmidt held forth on how much the company is doing to address the concerns that caused Viacom (VIA) to take Google to court over YouTube’s policies regarding content it doesn’t own. He said filtering technology is in “various stages of rollout,” as if that were good enough. He said Viacom “rushed” into litigation. He obviously has the luxury of knowing the suit has done nothing to blunt YouTube’s advance.

On the subject of social networking, and Facebook in particular, they made it very clear just how interested Google is in getting into the game in a more meaningful way than its Orkut service. (Read my colleague Josh Quittner for a contrary view on how much Facebook worries Google.)

Bottom line: From my perch, these guys were cool as cucumbers. Genuinely relaxed, engaged and at the top of their game.

By the way, Google reports earnings Thursday. The company’s worth $200 billion. And its founders give off the vibe that they’re just getting going. As I’ve written in the past, Google’s management isn’t perfect. Its previous quarter was sloppy. My gut tells me this one won’t be.

A technology generation ago Sun Microsystems (JAVA) extremely successfully convinced investors that it put the “dot into dot-com.” It was a brilliant marketing campaign. The product hadn’t changed, just the perception of it. (Ed Zander, whose best effort at Motorola (MOT) has been embracing the pre-existing and relatively empty “seamless mobility” label, likes to take credit for the marketing move at Sun, where he was the president.)

Now Sun is playing the perception game again. It will do a reverse 1-for-4 stock split in an effort to convince investors it is not a loser company. Executives acknowledged that reverse splits are meaningless from a valuation perspective. (Ditto for regular splits, by the way.) In fact, a reverse split usually is a sign of desperation. But Sun says it can stop the questions about the company’s staying power if essentially ignorant customers stop seeing a sub-$10 stock price.

Last week Sun also dropped its long-time stock symbol, SUNW, in favor of JAVA, which is the name of pioneering software Sun developed. (Google (GOOG) CEO Eric Schmidt played a key role on Java.) “More than a billion people across the globe, representing nearly every demographic, market and industry, rely upon Java’s security, innovation and value to connect them with opportunity,” CEO Jonathan Schwartz said in a statement. “That awareness positions Sun, and now our investor base, for the future.”

Whatever you call it, the stock was up 11 cents, or 2%, Thursday, to $5.48. Yawn.

Two thoughts ran through my head as I read a column by my most excellent colleague Geoffrey Colvin called Don’t go gaga over Google. The first, given that Geoff’s piece was all about why shares of Google (GOOG) are in no way worth more than $500, is that it would have been at least sporting to have mentioned that we are the magazine that not quite three years ago asked, on our cover, if Google really was worth, wait for it, $165 per share.

I’m not enough of a student of economic value added, or EVA, analysis that Geoff deploys in his piece to judge how conclusive his argument is. I’m also not sure how closely professional managers follow this method. I’d love to know confessed-Google-lover Bill Miller’s thoughts on the subject. Actually, I’m not really commenting one way or the other on the valuation. Not now, anyway. All great companies fetch a premium that defies any rational analysis — until they don’t. Witness Microsoft (MSFT), whose stock grew until 2001 and hasn’t since, and General Electric (GE), which, as the one and only Nelson Schwartz reminded us this past weekend, still hasn’t recovered its 2000 high. (Actually, Geoff Colvin wrote that story too, in 2005. But now I really digress.) In essence, Google’s valuation will remain tied more to its ability to grow than its return on capital. That’s my opinion, for what it’s worth.

Which leads to my second thought, regarding a line near the end of Geoff’s column that journalists often refer to as the “to be sure” line, as in “To be sure, Mr. Smith accomplished much in his career …” Geoff writes:

Irrational valuations can last a long time, and sometimes they correct gently rather than violently. And it doesn’t mean that Google is poorly run. On the contrary, it has been brilliantly run. (emphasis added.)

That’s a fascinating point, because here in Silicon Valley there’s absolutely no consensus that Google has been brilliantly run. There’s no question that Google has brilliantly exploited a massive opportunity in the online advertising market, primarily search-based text ads. No one can ever deny that.

Whether the company is well run, however, simply can’t be known yet. As I pointed out in my own cover story last year, Google so far has been able to avoid answering the question of whether its chaotic nature is by design or whether it’s merely holding onto the handles of one incredibly fast roller-coaster ride. Its young founders are universally believed to be really bright guys. But they’ve never worked anywhere else. They condone, nay, encourage, a permissive culture that lets engineers run wild whether or not they are contributing to the bottom line. Its CEO, Eric Schmidt, was a top scientist at Sun Microsystems (SUNW) and then an uneventful CEO of a relatively unimportant Novell (NOVL). The line on Schmidt is that he has grown tremendously in the job at Google. That’s undoubtedly true. Still, he hasn’t been tested by the kind of adversity that knocks CEOs on their backs. And to judge by one data point, Google’s surprising inability to manage its hiring costs, Schmidt hardly has the place running like a finely-tuned engine.

To be sure, Google already is a company for the ages. Its stock may surge even more. The greatness of its management, however, will be judged far more in the next three years than in the last three since it went public.