Who the hell is Bob? He is an engineer, an early developer of the Internet, an inventor of Ethernet, an entrepreneur and a critic. He once admitted his mistakes in public because he misjudged the development of the Internet, and now he has become a venture capitalist. For 40 years, bob metcalfe has been at the forefront of the engineering technology industry. Recently, he and Bai Wanning, executive editor of EE Times, talked about his investments in optical equipment, wireless/RF, hybrid fiber/coaxial and supercomputers in Polaris Ventures' office in Boston's high-tech corridor. In the meantime, Kan Kan talked about his personal views on the future of the Internet, the rise of blogs, the decline of print media, why engineers should not set foot in venture capital, and how to solve the energy crisis.
Bob metcalfe, inventor of Ethernet.
EE Times: What technologies can interest you at present?
Bob metcalfe: Everything related to video and positioning system excites me. The application of GPS is becoming more and more popular, so any company that produces, buys or processes geographic information is a big undertaking, which is also very exciting.
At the same time, I am also focusing on my first energy business. I am determined to rid the world of this chaotic energy state, but I don't want to be seen as an exclusive monopoly group or a madman who manipulates environmental protection organizations in other overseas countries.
EET: So, what's your better solution to the energy problem?
Metcalfe: Improving the utilization rate of gasoline, using nuclear energy, reducing pollution and adopting carbon sequestration technology will all help solve the energy problem.
EET: How many companies have you invested in?
Metcalfe: Five. The first companies are Narad, which provides network services of 100Mb/s on the fiber/coaxial hybrid network, Ember, which studies ZigBee, Paratech, designs tunable circuits of RF front-end, and SiCortex, a supercomputer company that designs Linux clusters. The fifth company is Mintera, which produces optical transmission equipment.
EET: What role do you think venture capitalists should play?
Metcalfe: Usually, venture capitalists take the time to choose companies and help them. At first, I thought that if a successful company (such as 3Com) was established, it would definitely help. From the past five years, I have learned that it is more important to choose a company worthy of investment than to provide its help. Of course, you must take care of both, but if you want to create greater returns for investors, which is also our job, then you must choose very carefully, because the help you can provide to a company is very limited.
EET: Have you ever thought that business will improve quickly? Has this really happened in your experience?
Metcalfe: During the Internet bubble,' quick U-turn' was once the law of the market, but now this phenomenon has rarely appeared. What we can plan at present is that there is a preparation period of 5-7 years from initial investment to capital flow.
EET: Why did you become a venture capitalist?
Metcalfe: I have been a publisher and critic in high-tech publishing house for 10 years, which is enough for me. I hope to stay in the innovation industry, but do something different; 10 is the time to do something new. Therefore, on the basis of my column about the ups and downs of the Internet in the past decade, I published my book "Internet Collapse and Other Information World Authorities". Then I started to do venture capital, because this is? ~ Let me stay in this innovative industry.
EET: What is the basis of your writing this book?
Metcalfe: I just compiled some columns about extreme counterexamples. The title of the book comes from a series of columns that predict that the Internet will collapse in 1996. In fact, I quantified the characteristics of the network crash, and my prediction was only eight times worse than the actual one, but I still ate the column because of my miscalculation. Metcalfe once smashed this column and water in a blender on the stage and drank it with a spoon. For many people, the fact that I ate that column at 1997 is more famous than the fact that I invented the Internet.
EET: What does this book say about the future of the Internet?
Metcalfe: In one of my columns, I talked about seven or eight reasons for the Internet crash. Safety is one of them, and I'm sure I'm right. Then there is spam.
EET: Is there a solution to spam?
Metcalfe: There are two ways, one is to use economics, and the other is to use licensing rights. These two methods are not within the original scope of the Internet, and we need to improve them today. Adding some economic information around e-mail can greatly reduce the emergence of spam; The method of using permission to restrict spam means that you can't send emails to the other party without permission. In addition, email should be strictly encrypted at all times, and no third party is allowed to read our email, so it is actually a mistake to adopt email screening program. Of course, the best way is to combine economy with permission.
EET: what about safety?
Metcalfe: I wrote the first RFC (draft for comment), describing the vulnerability of the network to security attacks. This is a warning because we forgot to add security functions to the Internet. I once suggested that anonymity should not be used as the initial setting method, but this access method is used in the Internet now. Source field packets should be checked routinely, but we don't actually do so. Of course, in some cases, there should be a way to achieve anonymous access, but it should not be used as an initial access method. This is why spam, viruses, worms and Dos (Denial of Service) attacks are rampant, all of which are attributed to the anonymous initial setting of the network.
EET: Has this problem been solved?
Metcalfe: Not yet. In order to solve this problem, the router vendor needs to start the function of checking the source field. I believe that someone has already done so. Routers like Juniper can check the source field, but no one has started using this feature yet.
EET: Why not use it?
Metcalfe: Maybe they are not as smart as me!
EET: Do you still think the Internet will collapse?
Metcalfe: I never said that the Internet would collapse. What I predict is a crash, which means a shutdown. I even quantified the extent of the crash: I predicted a kind of "gigalapse", that is, during an interrupted execution, 654.38+billion hours of user time was lost. 1996 The maximum interruption is 654.38+065.438+08. Although it was very close to Gigalapse, it still didn't come true, so I ate my column.
The Internet is still quite fragile, but it also has inherent resilience. Maybe my terrible warning has been exaggerated, and the Internet is getting better and better. In the next decade, we will develop video applications on the Internet, such as video mail, video conference, video on demand and video marketing. Our next generation has started downloading and copying CDs online, and soon they will secretly copy DVDs.
EET: What do you think is the most interesting development on the Internet today?
Metcalfe: I think blog is probably the most interesting phenomenon at present. I witnessed that blogs destroyed the old pattern of journalism, and the situation of daily newspapers went from bad to worse (culpable of punishment). Recently, The New York Times Boston Branch laid off 200 people. Because of the corruption in The Boston Globe and The New York Times, I am very grateful for this incident. What an amazing coincidence! The publisher in The New York Times happens to be the son of the former publisher! The current publisher happens to be an extreme leftist. He doesn't understand the value of newspapers at all, but only regards newspapers as a personal propaganda tool. However, readers understand this, so they don't support it as before. Therefore, when online blogging flourished, he fired a large number of employees. I think this symbolizes a bright future, because people can have more choices and greater freedom, and it also brings more competition and diversity.
EET: But the content of the blog is unedited.
Metcalfe: As time goes by, some blogs will start editing and filtering their content, which will enhance their reputation and diversity of content, and will produce more blog sites. I think the secret of internet progress is' freedom to choose competitive substitutes', which can be abbreviated as FOCACA.
We are all learning how to search and filter information. Google has been able to search web pages, and we are developing the ability to search blogs more effectively in order to link facts, opinions and information more effectively.
EET: At first, people hoped that the Internet could break down and eliminate communication barriers between different societies, but now some people think that it has spawned more closed groups, and a group of like-minded people only communicate with each other in this closed group. Don't you think this is just the opposite of people's initial expectations for the Internet?
Metcalfe: That's criticism. There is no doubt that the Internet spreads freedom all over the world. You see, everyone in the new york Times is in trouble. Of course, I'm not saying that newspapers should stop publishing, nor that everyone should cancel their subscriptions.
The printing industry will not suddenly die out, but it has indeed begun to decline. But it still takes a long time to pay attention. Although the display quality of computer monitors is getting better and better, don't forget that there are still four places where monitors are difficult to access, that is, beaches, bathrooms, buses and ... I forget what the fourth one is. These places can't take computers; But now, I think we can do it.
EET: Since you invented Ethernet in 1973, it has changed. How do you define today's Ethernet?
Metcalfe: Ethernet has changed a lot in its 33-year history. The word Ethernet is no longer its original meaning. I often say that the quality that Ethernet still maintains is its business model. This business model has six characteristics:
First of all, Ethernet is based on an industry standard; Secondly, compared with the open source code model, the establishment of this standard belongs to some companies; Thirdly, fierce competition among suppliers promotes development; Fourth, this competition is not based on incompatibility, because the market requires products from different manufacturers to be interoperable for users to choose; Fifth, the rapid development of standards is closely related to the interaction of the market; Sixth, no matter how fast the standard develops, its forward and backward compatibility is very valuable. This is the most time-tested part of Ethernet.
EET: Does Ethernet have anything to do with ZigBee that you are currently studying?
Metcalfe: Ethernet provides a solution for networking between different PCs; ZigBee is an embedded computer network solution. The earliest standard of Ethernet is IEEE 802.3, and the standard of ZigBee starts from IEEE 802. 15.4, both of which have their own standards. And both of them have top-level protocol stacks: TCP/IP for Ethernet and ZigBee for 802. 15.4. Just like the discussion about whether to adopt TCP/IP for Ethernet in those years, whether to adopt ZigBee for 802. 15.4 is controversial now. This issue is still under public discussion, but ZigBee has the highest voice. Others include Millennium net, crossbow, dust net, Zensys and the whole TinyOS series. ZigBee is just a commercially supported control standard.
EET: Does ZigBee have forward and backward compatibility like Ethernet?
Metcalfe: Yes. At present, Ember thinks that the standard 15.4 standard 2.4GHz radio frequency will be the most important ZigBee medium, but it is not the only one. In addition to 2.4G, there are magnetic, sub-Gig (a Hz) radio frequency and ultra-wideband (UWB), so the protocol stack must also be modified to adapt to various carriers under it. In addition, with more and more applications developed for ZigBee, more specifications must be adjusted. There is also a debate about whether to set IP (Internet Protocol) for monitoring on 802. 15.4 radio frequency. This is actually a good argument. If its technology hadn't developed so slowly, it might even have succeeded.
EET: What are the opportunities of ZigBee in the face of various challenges? How will it develop?
Metcalfe: If you look at the industrial control market that ZigBee is currently investing in, you will find that this industry, which is traditionally based on closed patents, has been severely divided. This market is difficult to enter, because they can't accept the concept that standards automatically establish such a concept; In fact, this has always been the direction of suppliers' efforts. Because they need to ensure immediate response, Ethernet is rejected by the industry because of communication conflicts. Because Ethernet can't provide such instant response, they don't buy it. Now you also hear a lot of similar arguments about ZigBee, such as "we want to use wired transmission, mesh things can't work, and the uncertainty is too great."
EET: Which communication medium will ZigBee choose among magnetic, ultra-wideband and 2.4GHz?
Metcalfe: I think all three will be adopted. According to our investigation, 2.4GHz has the greatest potential in the next few years.
EET: Which side are you on in the argument between optical transmission 100 Gbit and 40 Gbit?
Metcalfe: I must first declare that I am an investor in Mintera; Mintera is a company that sells 40-Gbit/s ultra-long distance optical transmission equipment. So in this debate, I am in a very special position. Manufacturers want to increase the transmission speed of their own networks by four times, that is, from 2.5 to 10Gbits/s to 40 GB its/s, but for Ethernet, we usually increase the transmission speed by 10 times, and now it has reached 10G. According to this logic, it should be upgraded to 100G next time. However, the practical principle cannot be ignored. Upgrading to 100 is much more difficult than upgrading to 40. So the question is whether it is worthwhile to upgrade to 100-Gbit Ethernet, how high its price will be, and so on. So we introduce and sell 40G equipment to manufacturers, and with this equipment, they can really transmit 3000km at 40G speed. We just proved it. Ethernet usually only needs to transmit 10 feet, so maybe you can achieve the speed of 100G on Ethernet. Of course, because we all use IP, and data packets may be transmitted at the speed of 100G in one place, but only 40G in another place, which makes different transmission speeds appear in the network at the same time.
EET: Has 40 Gigabit Ethernet regained its development momentum?
Metcalfe: Mintera has been running well, so it has survived the overall decline of the communications industry. At present, the company has cut off the failed business and entered a dormant period. A few months ago, we decided to invest again, because we felt that the network business was still growing, and the supplier's replenishment potential for optical fiber was still considerable. So although we are not sure when, it is expected that within three years, the market interest in 40G will rise again. People's understanding of 40G has begun to increase, which is evidence.
EET: Do you have any advice for those who want to be venture capitalists?
Metcalfe: Yes, they shouldn't be in this business. There are too many people engaged in this field now. At present, there are thousands of companies in this field, and the capital investment is countless. It's best to work in medicine, or be a blogger, artist or doctor, but don't do venture capital, because it's my job.
EET: Do you think you are still the outgoing Bob? Or do you realize that you are much more mature and stable?
Metcalfe: If anything, I have become more conceited. Like many people, I feel sorry for the current standard of speech, but as long as people I don't agree with are always disgusting, I will definitely be rude to them.
Introduction to robert metcalf:
Born on March 7th, 1946, birthplace: Brooklyn, new york.
Education: 1996 Bachelor of Electronic Engineering and Bachelor of Management from Massachusetts Institute of Technology; 1970 Master of Applied Mathematics, Harvard University; 1973 Doctor of Computer Science
Work experience:
200 1 1 Up to now: Partner of Polaris Venture Capital, specializing in Boston IT emerging companies.
1990 ~ 2000 (CEO of1990-95): I wrote an internet column in InfoWorld for 8 years, with more than 500,000 readers per week.
1979 ~ 1990: 3Com company was established, and since then he has served as president, CEO, general manager and vice president of engineering, sales and marketing department in different periods.
1972 ~ 1979: researcher of Palo Alto Research Center (PARC), majoring in computer science, and then studying system architecture management.
1965 rpm 1972: engineering research of Raytheon, Adams Society, MIT Nuclear Science Laboratory and MIT Mac project.
Works:
Packet Communication (published by Thomson), Internet Collapse and Other Views on the Information World (published by IDG Books), The Next 50 Years of Computer Computing Technology (co-edited with springer Publishing House, etc.).
Recent awards:
2005: He was awarded the National Science and Technology Award by President Bush for inventing Ethernet.
In 2003, he won the Marconi International Research Fund Award and was awarded three honorary doctorates.