Archive Magazine 1996

home *** CD-ROM | disk | FTP | other *** search

/ Archive Magazine 1996 / ARCHIVE_96.iso / discs / mag_discs / volume_9 / issue_12 / clippings / Hauser < prev next >

Wrap

Text File | 1996-07-30 | 21KB | 261 lines

<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML//EN"> <html> <head> <title>Hermann Hauser</title> <meta name="GENERATOR" content="Microsoft FrontPage 1.1"> </head> <body bgcolor="#FBFFE6"> <img src="../gifs/interview/fryuphd.gif" alt="The Friday Fry-up" align=bottom width=393 height=56> <hr size=1 align=left> <div align=left> <table cellpadding=7 cellspacing=7 width=100%> <tr><td valign=top><tt>Friday, July 19</tt><tt>And Bill said: "What's a network?" </tt>—<tt> Acorn's Hermann Hauser</tt> Acorn's Hermann Hauser on the NC and getting it right this time ... Hermann Hauser has probably been upbeat since the day he was born in Vienna, Austria — you don't foster 20 technology start-ups without some sort of optimistic streak. But right now, his confidence knows no bounds. Acorn, the company he founded in 1978 during post-doctoral work in Cambridge is about to deliver the NetStation, the technological and market prototype for the Network Computer Alliance founded by Oracle, IBM, Sun, Netscape and Apple. Acorn lost millions last year — yet on the day in December when it made a joint announcement on the NC with Oracle, its stock nearly doubled in value. Since then, it has jumped, as Hauser notes drily, "every time we make an announcement". US investors have persuaded former majority holder Olivetti and Hauser himself to part with stock. The NetStation runs on ARM chips made by Adavanced Risc Machines, in which Acorn has a 43% holding. The company's processors are about to go into 3DO platforms, GSM phones and auto engines. Things are going well — and Hauser puts it all down to the forging of alliances. <tt>What brings you here? An evangelism tour for the NC? </tt> An evangelism tour combined with seeing my relatives in Warkworth. I married a New Zealander and my wife and kids are spending part of the British summer here. <tt>I have to say I was impressed by the NetStation prototype, but it did occur to me that Acorn has been first many times before. Do you lie awake at night worrying that you'll be first again , but that someone else will make the money?</tt> No I don't, because I learned why we didn't capitalise on our previous firsts as much as we should have — and the reason was alliances and partnerships. I often tell the story that Bill gates was trying to sell me MS-DOS in the early 80s and I had to say "Bill, we can't possibly take such a retrograde step, because our operating system really is an operating system and has many features that MS-DOS doesn't have. And by the way, a little schoolboy can type 'I am Johnny' into one of our computers and be logged on through the network to a local fileserver. They can use the same commands to get files down from the server that they've learned with a floppy disk." And Bill's answer to that was, "What's a network?" So they were clearly far, far behind where we at the time. So the question is why is the world not Acorn compatible instead of Microsoft and Intel compatible? And the answer is partnerships. He picked his partners much better than we did. In fact, we didn't even think about partners because we were so successful. At the time, we were the fastest growing company ever in the UK and we didn't think we needed partners. Neither did anybody else. Bill just had a bit of Bill luck and got the right partner at the right time. But now we know that it is all about partnerships and so for the Network Computer we have the world's second-largest software company, Oracle, as the lead partner. It was Larry Ellison's vision of the Network Computer that just fit like a glove with the Acorn NetStation. Together we now have some 50 of the biggest IT companies in the world signed up to this standard. So I'm as confident as one can be at this stage of proceedings that the NC will be a success. <tt>I'm interested that Oracle seems to have come to you for a hardware solution, because in all of this, amid the talk of the ARM chips and Acorn's ability to build the box, I don't hear much about the RiscOS.</tt> You are mistaken. The Acorn NetStation runs the Acorn RiscOS. It's now called NCOS, but it's just renamed RiscOS. <tt>So all those licensees will be running the RiscOS?</tt> Yes. Absolutely. <tt>That's quite extraordinary. That's a very fast trip from a niche market to the mainstream.</tt> Absolutely right. <tt>Quite apart from the technology, a crucial part of making a success of the NC would seem to be selling it at the right price.</tt> Yes — £399, including VAT, in the UK. The reason we can achieve that price point is well-known — it's the ARM chip. In this case the ARM7500fe, which is basically a single-chip computer. On the chip you have the microprocessor, the I/O controller, the video controller, the DRAM controller. And it costs less than $30. If you were going to do that in the PC world you'd have to find a processor somewhere between the 486 and the Pentium — that alone will set you back about $50 and then you've got to buy the rest of the chipset. Once you've got that together you're way above where we are with the single chip. That's reason number one. Reason number two is RiscOS. We have put it on ROM for more than 10 years, and as you probably know, operating systems don't like to be ROM'd. It's very hard to do, because if you have a bug in your operating system, it also gets ROM'd. But because we've done that for so long, we think we know how to do it better than anybody else. And ROMming it is key for ease of use. When you switch this thing on, it's ready faster than the TV warms up. And ROM is a fifth the price of RAM and there is no room to have a storage device, you can't afford to have a disk in there. So it must be in ROM — and you can't do that with Windows 95. The third reason is the quality of the picture, using anti-aliased outline fonts — something else Microsoft is still trying to get right in Windows 95. We got that right five years ago. <tt>Bill Gates put out a white paper on the NC recently — and one of the things he was most scathing about was having the OS in ROM. He said it was fatally flawed because it couldn't be updated. What would be your response to that?</tt> Bullshit <tt>Can you expand on that for me?</tt> Well, I've gone through the advantages. The disadvantage of course is that you can't change it — or so Bill Gates would have you believe. Of course this isn't true either. Because we've ROM'd things before, we've worked out mechanisms for patching the OS in RAM. What Bill also doesn't tell you is that one of the standard slots in the NetStation is a ROM slot, so if you want to do a complete upgrade, and have it done by the user, we have a thing that's very similar to a PC Card which just plugs in and there's your upgrade. It's a cheaper, more reliable upgrade. Of course you'd only do that, say, once a year or two. All the minor updates sit on the network — the whole point of having a network computer is that the updates sit somewhere on the server and get downloaded without you even knowing. Apart from anything else, it's also highly resistant to virus attack. You can't alter a ROM over a network. <tt>The NC does seem likely give a whole new meaning to the "service" function of an Internet Service Provider. Suddenly the ISP has to provide remote storage, software maintenance and upgrades and so on. Are there service providers in the UK who understand what's required and are prepared to do it?</tt> They're falling over themselves. Every service provider we've talked to so far — including New Zealand Telecom — has been very happy at the idea of working with us. They're delighted with the extra service requirements, because it brings in extra revenue. <tt>Did you talk any sort of time frame with New Zealand Telecom?</tt> No, but I wouldn't be at all surprised if it was there early next year. <tt>What is NChannel?</tt> It is the company we have established to distribute the NetStation boxes. It's a spin-off from Acorn, to give us a presence in the consumer electronics market, which we didn't have before. This is a consumer electronic business — it's producing something that, as the Californians say, is a great out-of-box experience. You take it of the box, plug it into the power supply for the electrons, the telephone network for the bits, and the back of the TV set for the photons, and there you are, you're on the Internet. The aspect to ease of use is that you've got to make the Internet itself easy to use. So you've got to have a service component which gives people a satisfying experience the first time they go on the Internet. That's a very important part of the solution and for that we'll need partners around the globe. <tt>Ah ...I confess, I saw "NChannel" and thought, "Ah, they've got content too!" It does seem that the successful NC vendor may be the one which almost gives away its box in order to capture viewers for the content that will come through it. Sony certainly appear to be preparing to take that path.</tt> That's exactly the strategy. You've got to have a compelling offering which is low-cost enough to hit the consumer. And if that means subsidising the box for a while then that's what you do. <tt>So who's going to subsidise your box?</tt> NChannel. <tt>Who's putting up the money for NChannel?</tt> It's a start-up with venture capital finance and I put some of my own money in there because I feel this is a big opportunity. <tt>And you recently released some of your own stock in Acorn itself?</tt> Yes, but I have 20 companies that I have started over the years that are now worth some three quarters of a billion US dollars. These are all high technology companies and the role that I typically play is to put the team together, provide seed finance and see them through their early growth before sending them off to fight on their own in the big wide world. <tt>Most of those are British companied. Acorn, more than most has always been a example of British technological prowess and values, and that of Cambridge more so.</tt> Yes, very much so. You know, I always say that Cambridge until recently was a disaster area, because it had the lowest revenue per bright person in the western world. The revenue's not that bad — there are about 700 little hi-tech companies around Cambridge University with a combined turnover of about two billion pounds. There’s potential that we have there because of the quality of the people who go through the university — who are just as good as anyone at Berkley or Stanford or Harvard. But we're not translating the brilliance of these technical people into revenue as efficiently as they do in Silicon Valley. There are still two things that are missing from making Cambridge a European Silicon Valley. One is a better balance between technology and marketing and sales. The second thing is links with the US and Silicon Valley in particular. I'm spending most of this summer there and I'm there once every two months year-round. Whether we like it or not the IT agenda is being defined in Silicon Valley and we just have to learn to fit in with it. And if you're not willing to do that, you're not a player. I think the Acorn NetStation is a good example of that. It was clear that a number of big companies were ready to play in this field, and that Oracle was going to be the lead partner, and I'm very pleased that we've managed to hook up with them. Apart from the five founding partners of the NC Alliance — Apple, IBM,Sun, Netscape and Oracle — there are around 50 alliance partners. People like Mitsubishi, Funai, Olivetti, British Telecom, Deutsche Telekom ... lot of alliances. <tt>I'm interested that you talk about the Acorn NetStation as the NC, rather than an NC. Is it really that much of a fait accompli that Acorn's design will become the standard? </tt> There are two mantles that the Acorn NetStation has. It is the first instantiation of the NC standard But the Acorn NetStation is also the reference design that gets licensed to everybody else. The Acorn NetStation isn't exclusive to Acorn; it's just the first example of a Network Computer that gets licensed to Oracle or anyone else who wants to build a Network Computer. There are 50 other companies who want to do that. And that's why we started NChannel to make sure that there is also the Internet service part that makes the Internet easy to use. The only way this thing is going to take off is if the end user gets a compelling experience. <tt>There are two schools of thought about who the NC end user is initially going to be. One says it will be the corporate market, with users operating Java-enabled NCs on a LAN — and IBM's AS/400 "thin client" is the first expression of that. The other says it will be the consumer market. And you're picking ...</tt> Consumer. <tt>Why?</tt> Because there are more of them. We've done the market research and the results are just phenomenal. Whichever age and socio-economic category you ask — at £399 in the UK, $399 in the US, falling to £299 next year and £199 for Christmas 1997 — people want to buy it. You ask the people with PCs and, suprisingly, you get the same answer as people without PCs. It has very wide appeal and the reason is very simple. In London 60% of the billboard advertising now sports a WWW address. You can't open a newspaper without finding an article about the Internet. The awareness of the Internet is now 90%, but the number of people on it is tiny — so there's there are all these people feeling that something is going on there in the world and they realise they're not part of it, they feel excluded from what's happening. And they feel disadvantaged. With the Acorn NetStation we give people an easy to use and affordable box so they too can become part of that Internet community. For this coming Christmas, this and DVD will be the two consumer items people will buy. And DVD is quite a bit more expensive. <tt>At the risk of repeating myself to readers, I've always thought of the NC as the device which will let me send email to my Mum. She can type and hit a "send" button, but I don't expect her buy and to learn to use a PC.</tt> That's exactly the argument. My mother is now 74 years old and I've always wanted to give her one of my products. But I've never dared to because I know she wouldn't know what to do with it. This is the first product that I will give her for Christmas. She knows how to type, she loves writing and we've got a little video conferencing add-on for the NetStation so she can be in Vienna and see her grandchildren in the UK for the cost of a local phone call. <tt>That raises the bandwidth issue. For the foreseeable future it's going to be plain old telephone lines. Is that a limiting factor?</tt> Very much so. I think people will want more bandwidth the way they wanted more processing power in the past. The StrongARM chip has the ability to decode MPEG-2 streams at up to 15Mbits/sec in software. So that's up to an HDTV screen. That's the kind of capability you want, but at the moment you can't deliver 15Mbits/sec down a telephone line. But another one of my companies, ATM Limited, has just announced a joint development with Alcatel to transmit ATM cells down an ADSL link. So you keep the same copper in the ground but you have an ADSL link — at up to 51Mbits/sec. But in the short term, yes, there will be limits on bandwidth. <tt>For all your optimism about the NetStation and the ARM processors, the fact remains you're up against the two largest vested interests in the computer industry — Intel and Microsoft. How do you get around that?</tt> First of all, by having no fear. We have survived very well outside the Intel and Microsoft camp since our inception, so we know the bastards. And we fear them not. So they have locked up — some people would say stifled — the PC industry. They've been the main reason why the PC industry is so backward. It has one of the poorest microprocessor architectures in the world. It is a little-known fact that the Pentium is actually a very poor microprocessor. It happens to be the world's most popular microprocessor, but that doesn't change the fact that the architecture is rubbish. The ARM, for example is much cleaner and delivers a lot more mips per watt and per dollar, but that doesn't matter if you're non-standard. You can run a StrongARM off the heatsink of a Pentium and still deliver more mips than the Pentium which generates the heat in the first place. It's the same with Microsoft. Nobody will dispute the fact the Windows 95 is the best Microsoft operating system there has ever been, but if you compare it with the Mac environment or RISC OS it is clearly inferior. But there was a very memorable talk at Agenda 96 in Phoenix, by a guy called Robert McNamee, entitled 'Why Intel and Microsoft Don't Matter Any More'. The title was a bit premature, but it was a nice way to get people's attention. The point was that neither Intel or Microsoft owns the Internet. And this alliance around the NC shows that this is an opportunity to restructure the industry and bring together people from different industries and different regions. Sure, Microsoft and Intel will come up with a significant response to that in due course, but in the meantime we will have captured a substantial slice of the market. The NC is the fourth wave of the computer industry. The first wave was the mainframe, the second was the minicomputer and the third was PCs. If you look at the relationship between them, you find that the new one is always about half the price of the cheapest product of the old one and it is much easier to use and therefore addresses a much wider market. The minicomputer did a lot more than a PC, but the PC was a lot cheaper and it could be looked after by a single person. You'd didn't need a team of engineers to support it. A lot more sold. And people in America are now talking about a billion NCs being sold in the next 10 to 15 years.