You can't spend much time on the cutting edge of the computer industry these days without running into Hermann Hauser. The excitable 47-year-old Austrian with the world-class nose for technology seems to be everywhere. In person you'll either see him rushing around Cambridge (the English one) looking after companies or networking at Silicon Valley conferences. But it's in the world of ideas and corporate positions that his ubiquity really shows.
Put two geeks together these days and the chances are they're talking about Oracle's Larry Ellison and his vision of a $500 Internet computer that'll knock down the PC standard. Which means that, know it or not, they're talking about Hauser. The groovy little black box that Ellison has been waving around is made by Acorn, the computer company Hauser founded nearly 20 years ago. The speedy RISC chip that lets it run as fast as a PC while costing as little as VCR is made by Advanced RISC Machines (ARM). That's Hauser, too - the firm started as an Acorn subsidiary, and was then spun off in a deal with Apple.
British homes with network computers will dial into the Internet through access providers such as Hauser's UK Online. Their data streams will flow through telecoms networks that use ultra-fast Asynchronous Transfer Mode switches like the ones made by Advanced Telecommunications Modules, Ltd (ATML), which Hauser founded three years ago. Those streams will offer commerce and entertainment provided by companies like Electronic Share Information, an online stock market service, and Entertainment Online, a games and education venture - both Hauser productions. It goes on. Someday the window on to the Net may be a flat screen like those being developed by Cambridge Display Technology (founding director: H. Hauser). Or maybe tomorrow's couch potatoes will simply use their existing televisions, turbo-charged by smart set-top boxes made by Online Media, an Acorn spin-off. Some might even use network computers that look like books, such as those developed by Olivetti's Cambridge Research Laboratory, started by Hauser when he was at Olivetti and run by his long-time business partner, Andy Hopper. It could be that AI technology from Harlequin will make the boxes smart; or that ray-tracing chips being developed by Advanced Rendering Technology will give them photorealistic graphics capabilities. If so, Hauser wins.
He has a finger in virtually every important technological pie, and he has baked quite a few of them himself. He bet on cheap networking early, and the world came his way, just as he knew it would. From his desk in Mount Pleasant House - his Cambridge headquarters and home to four of his companies - the promising view looks all the better for being long expected.
But before surmising that Hauser is looking at a sure thing, there's something you should bear in mind. Not only can he miss: he usually does. His failure in the market is almost as legendary as his technological insight. Of the dozen computerrelated firms he has had a hand in starting, only two are still in business and profitable, as best I can tell (most of the firms are private and do not release financial results). True, in the world of start-ups, profitability is a tough test - most such firms spend years losing money while they develop their technology. But after nearly 20 years in the industry, Hauser is still waiting for a hit. Despite their promise, none of Hauser's firms have yet made it out of technological or national niches.
To expect a Cambridge entrepreneur to start another Microsoft might seem to be asking too much. But Hauser sets himself up for such a comparison. Though every entrepreneur wants success, Hauser wants a specific sort of success: a very big one founded in a particular place. America's high-tech hegemony upsets him. "Having been born in Europe, I think it's a worthwhile place to live. But at the moment, Europe is doing its best not to survive." His aim is more than just personal wealth (which he already has) - it's a European technological renaissance. And he's chosen Cambridge to be his Florence.
Hauser came to Cambridge as a teenager studying English and fell in love with the place. In the 1970s he returned to King's College to do a PhD, and never left. He loved the town's research talent, and he saw its potential for entrepreneurial drive. As yet, however, that drive hasn't got very far. The mid-1980s "Cambridge Phenomenon" generated a lot of small technology start-ups, but few got much bigger. Without America's big domestic market, venture capital supplies and enticing IPO environment, the town's high-tech industry has shown a marked failure to thrive. "At the moment," says Hauser, "Cambridge is contributing bugger-all to Britain's GNP."
Which only makes him more obsessive about seeing the town give birth to a US$1 billion high-tech company. It doesn't matter all that much which firm scales the $1-billion heights; it's not even that a $1-billion firm is such a good thing in itself. What matters is creating an environment where such a thing is possible - and seen to be possible. Under such a tree many acorns, so to speak, might grow.
History lessons
If it happens, it will be because of Hauser. He has helped start some 20 firms (even he loses track of the exact count); he is the biggest individual investor in Cambridge's high-tech industry. He is also its ambassador at conferences and meetings. "Where I see an overlap between a Cambridge technology and a trend, I bring it back."Any ambassador makes connections; Hauser has made a lot of good ones. He's not exactly a master of the Californian hard sell à la Steve Jobs: he's too soft-spoken, too Austrian, too much the sort of person found as a dining fellow at Corpus Christi College. But he can pitch technology vision with the best of them, and that wins him allies. His deal with Oracle is just the latest such arrangement; he's had a long-standing alliance with Apple and has founded several companies with Olivetti.
Such success is even more impressive set against the record of his past. Hauser's firms may be worth a lot - he puts the figure at about US$500 million - but they are also collectively a long way into the red. Acorn, his first company and biggest success, lost more than £12 million last year, four times its loss of the previous year. This isn't directly Hauser's fault - he hasn't run the company since 1985. But Acorn has never climbed out of the hole Hauser left it in when, one Christmas, he made 250,000 computers that nobody wanted to buy.
Acorn's computer was clearly better than the competition, and Hauser has always operated on the principle that, given a choice, consumers want the best for their money. When Bill Gates tried to sell him MS-DOS in the early 1980s, Hauser sent him away empty handed; why would he want to license a clearly inferior operating system? In the mid-1980s, Hauser says, "We were looking at our UK market share and feeling pretty good about it. What we didn't realise was that it was a global game." The consumer knew better. Though muscular American marketing helped, mostly the consumer wanted a safe standard. When the IBM PC took off, Acorn's market crumbled; Hauser wasn't the sort of guy to make clone machines. Olivetti stepped in to buy up the pieces.
Watching Acorn choke "was very sad," Hauser says. "We had great expectations for it." But it didn't trouble him too much to lose day-to-day control. Hauser is no manager and he knows it. Today he runs none of his companies - he's strictly the non-executive director, helping to steer the strategy. Somebody else has to get the forecasting right and meet the payroll. Olivetti agreed with Hauser's self-assessment. It stopped him running Acorn but hired him as vice president for research. In that role he set up research labs around the world, including the one in Cambridge for his old pal Hopper. The consequence: some terrific labs, but a terrible computer company. Olivetti has lost more than $2.5 billion since 1990, and is now considering getting out of the PC business.
In 1988, after a stint in Italy, Hauser came back to Cambridge. This time he was swept away by the possibility of realising something like the legendary DynaBook - Alan Kay's 1970s concept of a computer as easy to use and as natural to hold as a Tom Clancy hardback. But after five years of development Hauser and his company produced the EO personal digital assistant, which cost more than $1,000 and recognised handwriting even worse than Apple's much ridiculed Newton. It sold fewer than 10,000 units before AT&T, which had bought the company in part to market it better, put a merciful pillow over its little face. EO's backers lost nearly $100 million. Even ARM, whose RISC chip technology is widely acclaimed, has not been the worldbeater its backers had hoped. Spun off from Acorn to allow Apple to take a stake, ARM made the Newton's processor and looked forward to a fortune when Apple's John Scully rashly promised that personal digital assistants would create a "trillion-dollar" market. The ARM chips are cheap and fast, and consume so little power that - as the company demonstrated in a neat stunt with a thermocouple - you can run one off a Pentium's waste heat. But the fact that, pace John Scully, the PDA market is currently worth more or less nothing means that today most ARM chips are found doing rather less wired jobs such as running electronic chess games and automatic braking systems in cars. The digital mobile-phone market for them is growing fast - but there's not yet a billion dollars to be made in it for a small chip designer.
Scores on the doors
Hauser was concentrating on the future before his past looked messy; he's not about to stop now. He may have been wrong about the evolution of the PC industry, personal digital assistants and the value of standards over bithead fantasies - but he was right about networking. The pride in his boast that "Acorn never sold a computer that didn't have a networking connection" is deserved. In the Cambridge computer lab where Hauser got his start, Andy Hopper and his colleagues developed the Cambridge Fast Ring network that led to ATM switches. Acorn's networking technology was the model for Apple's AppleTalk.Hauser didn't anticipate the explosive growth of the Internet - who did? - but when it happened, it proved right everything he has been saying for the past 15 years. The Internet offers a way to put Acorn into the mainstream. It is Hermann Hauser's second chance. The Internet's open standards mean that any sort of computer can join. This levels the playing field, eroding the advantage of standard setters. In the new network computer vision touted by Ellison (and Sun, and others) a cheap Acorn box is every bit as good as a Pentium PC - or even better, because it isn't lumbered with the millions of lines of code and thousands of transistors sacrificed to the gods of backwards compatibility.
Rather than starting with a hard drive creaking under a full load of applications software, half of which (ah, but which half?) will never be used, a network computer will basically run just one program - a Java-enabled Web browser which allows it to download and run any applet written in Sun's Java programming language. These applets are small software modules dedicated to single jobs. Users aren't forced to load and run a 20Mb spreadsheet for even the simplest computational task. Bells and whistles can be grabbed on an if-and-when basis.
Acorn's computers flopped in the marketplace because they didn't run the most popular applications and games. If tomorrow's hot applications and games are Java programs, and the killer apps are email and the Web, it hardly matters what kind of box sits on your desk, or in your front room. So Acorn has a chance. But what sort of chance?
It's easy to see why people like Ellison and Sun's Scott McNealy are so excited by this vision: if the centre of the computing world moves from desktop to network, the hot machines will become the network servers that store and deliver all those Java programs. Hauser's interest, though, is a little harder to grasp. If, as Sun's mantra has it, the network is the computer, the computer becomes not much more than the keyboard and screen. You want one that's cheap, performs adequately and looks good. Strong brand, a bit of design and low margins would seem to be the name of the game, just as in so many minimally differentiat- ed mass-market consumer goods. In other words, just when the desktop opens up again in technology, it closes in marketing. Nintendo, Sega, Sony, Philips, even the Power Ranger's Bandai are all coming out with network computers, for about the same £299 that Hauser reckons the Acorn network computer will cost.
Hauser reckons that the Acorn has three technical advantages over the games manufacturers and the PC makers in the battle for the newly standard-free desktop. The ARM chips offer more bang for the buck. Acorn's multitasking operating system is small (it runs in one megabyte of RAM) and can be permanently burned into read-only memory (ROM) chips, potentially eliminating the need for a relatively expensive hard drive. And Acorn's original computers were designed to use a TV as a monitor. Getting good images on a TV that has less than half as many scan lines as a computer monitor requires fancy optical illusions - in particular, a trick called anti-aliasing, in which text is actually blurred to make it more readable. Acorn has learned to produce such text in any font and size on the fly. If price really matters, then people will want to use TVs as monitors, and Acorn will have an edge.
Hauser is in earnest, but it's hard to take these advantages too seriously. It may not be easy to put an operating system in ROM or generate fuzzy text, but the games machines already do it, albeit not as well as Acorn does. If Sony wants to do it better, then I'd bet it can develop - or buy - everything Acorn has and more in a matter of months. And as good as the ARM chips are, some of the games machines have even better ones.
Whoever wins in the market, though, there's still the question of how big it will grow - and how quickly. The weakness of network computers is that they are only as fast as the network to which they connect. As the keen gadget freaks who buy the first network computers this year will discover, the Internet is not ready for couch potatoes. At 14.4kbps, Web pages dribble in even more slowly than teletext. How long will the average punter be willing to stare at "Host Contacted. Waiting for reply"? Even at 28.8k (the speed of next year's crop), anyone actually trying to run Java programs off the network will quickly remember why the PC beat the mainframe way back when. Until real bandwidth to the home becomes a reality - probably through upgraded cable TV networks using ATM switching to deliver up to 10 million bps - the network computer seems destined to stay in PDA land. It's hard to believe that Acorn, repurposing its 1980s home-computer technology, will be able to keep its first-mover lead during a long technological phoney war.
That may be bad news for Acorn, but Hauser has plenty of other irons in the fire. Take those ATM switches in the cable TV networks. ATM is starting to take off with the telecoms and cable companies because it is so well adapted to multimedia. Packets of information going from a to b across today's Internet are of many different sizes and will take many different routes. That means they'll often arrive in the wrong order, and sometimes not at all. If those packets are meant to be a continuous piece of video or music, that's bad: Stravinski can come out sound- ing like Skinny Puppy.
ATM solves this problem by using a small, fixed packet size. This lets ATM switches do most of their processing in fast hardware, rather than slow software. The result: 155 million bits per second, digital crack for bandwidth addicts. Hauser's friend Hopper is a good example of an ATM nut; his office has cameras and microphones on every available surface, all streaming video of him back through desktop ATM switches to workstations throughout the lab. ATML, the company he founded with Hauser, makes those switches. The sooner the world starts buying and deploying them, the happier (and richer) both of them will be.
But ATM is not taking off quite as quickly as its enthusiasts had hoped. The problem is that ATM is still expensive and largely incompatible with the existing Ethernet standard. The Internet took off because it could just colonise existing telecoms networks. People could experiment with Internet applications without a big investment; once they found some truly compelling use for the thing (say, the Web), they could invest incrementally in the network to make it better and faster. ATM, on the other hand, usually requires users to make a leap of faith: replace your cabling, your software and your switching, and you'll be able to do cool stuff like desktop videoconferencing. At the moment, big companies show precious little appetite for desktop videoconferencing or for other high-bandwidth desktop multimedia. What demand there is can be satisfied by upgrades to existing Ethernet networks.
No problem: Hauser's got other irons. There's UK Online, for example. This is an Internet-based online service, a bit like CompuServe or AOL, but specifically tailored for the UK family: local content, no porn. Its backers promote its pedigree - it uses the UK SuperJanet academic network as its backbone, and Tim King, its managing director, is a UK networking god. For keen Hauser watchers, this should set off warning bells; it's the old problem of cool technology over market focus. Customers don't care about the backbone, they care about the content.
Sure enough, UK Online has had a slow start since it went live last autumn. The company isn't releasing exact current subscriber figures; King says he hopes to have 25,000 by the end of the year. Hauser says there are a few thousand now. Part of the problem is advertising - there hasn't been any. Another problem is content, or lack thereof. Earlier this year, Jennifer and Roland Perry, its marketing and business development directors, quit in frustration over the lack of content development. For a while, the only staff member putting together editorial material was King's personal assistant, although Hauser says that the company now has six employees working on content.
Electronic Share Information, though, is doing better and matters more - its aims are particularly close to Hauser's heart. Its launch got the sort of publicity that money can't buy, thanks to a decision by the London Stock Exchange (LSE) not to provide realtime quotes on company shares for the service just days before launch. As the writs flew, the battle hit the headlines. Hauser claimed that the LSE was exploiting its position to restrict competition; the LSE said he had breached a non-disclosure clause in his contract. In the end, the exchange caved in: ESI was up and running with a live feed by December.
Real-time stock quotes are a start, but Hauser has much bigger plans for ESI. One of the reasons the Cambridge Phenomenon fizzled, he says, was that it was hard for young firms to raise enough money early on. In America, high-tech start-ups have a choice of venture capital firms eager to bankroll them. The venture capitalists, in turn, know they can get their money back by taking the company public early on NASDAQ, the low-cost small-company alternative to the New York Stock Exchange.
Until recently, Britain has had no NASDAQ. The LSE's costs quickly eat up the gains of going public for smaller firms. This, in turn, limits the venture-capital market - investors see no quick way to cash out. The consequence is that most British high-tech firms have had to grow organically, on their own sales - a poor prospect in fast-growing markets where early money is the key to expansion.
Hauser sees the Internet as a way out of this. NASDAQ, like new small-capitalisation markets in other countries, lets buyers and sellers contact each other directly. Removing the market-making middlemen who dominate big exchanges lowers costs by a factor of five. The remaining costs are accounted for in part by proprietary networks, which is why Hauser thinks that a market on the Internet could be cheaper still. He reckons ESI's costs will be less than a tenth those of the LSE.
Cost isn't the only advantage. The Internet is available to all, so investors can trade directly without going through a broker. And it is interactive, which raises the prospect of share auctions. For low-volume shares, where one cannot always match up a buyer and seller, ESI could schedule an auction. Sellers quote the lowest price they'll accept, and buyers the highest they'll pay. As the auction approaches, each revises his bid as he sees how others are acting. In the end, Hauser says, drawing a chart, the market settles on the price that generates the largest number of trades.
Originally, Hauser had planned ESI as a Cambridge-only market. After all, it's Cambridge that he thinks needs it most. But later this year he plans to launch it across Britain. ESI is also talking to firms in six other countries about setting up similar markets as ESI franchises. Eventually, such markets could be linked, allowing investors to trade from their desks in shares across Europe.
A few years ago, this idea would have been irresistible, had it not been unthinkable. But now that such things have become possible, it faces a lot of competition. Last year the LSE set up AIM, its own small-capitalisation market. Similar new markets are now running in France, with other countries soon to follow. ESI may be able to do it more cheaply using the Internet, but it pays a penalty in consumer confidence. Right or wrong, people still think of the Internet as a hackers' paradise, and the evidence so far suggests it will be a few years more before many feel comfortable investing their savings online.
Where does that leave Hauser? Right where he likes it: with lots of opportunity, plenty of balls in play and new technologies exploding all around him. Hauser claims that he - and other Cambridge firms - have learned from the mistakes of the past. Now, he says, the boffins bring good marketing, management and sales talent in early: "You wouldn't believe that this is the Cambridge of ten years ago." But watching Hauser's current ventures at work - how can you launch a nationwide family service like UK Online and not advertise? - it's hard to see the change.
Personally, Hauser will do fine. He is, after all, a success by many standards - just not by his own. When Acorn went public in 1983, he spent a brief spell as the twelfth-richest man in Britain, and before Acorn crashed he wisely banked a few million shares. His investments in small companies since then, he says, have made a twenty-fold return. Even EO did not hurt him personally - AT&T may have lost $40 million after it bought the company, but Hauser made money when he sold it. And what's good for him is good for Cambridge. In Silicon Valley, start-ups turn to "business angels": 40-something multi-millionaires who made their pile in the first high-tech wave, burned out and now invest as a hobby. Hauser is more hands-on than that, but he serves the same function - "call Hermann" is regular advice for Cambridge's would-be entrepreneurs.
But it's hard to imagine that Hauser will ever again be in the position to turn down a begging Bill Gates. Not that you never get a second chance - Steve Jobs is on his third, and a paper billionaire. But for all the new chances the Internet offers, one old rule remains: technology doesn't win; companies do. That's why Hauser's dream of a giant matters, and that's why it's still fair to question his ability to bring it about. Just because Larry Ellison is waving your prototype around doesn't mean that you're going to get rich. Especially if he's just come back from Asia negotiating deals with consumer electronics firms to make the real thing.
Jackie Bennion is a senior copy editor at Wired US.