Friday, November 23, 2007

Profile: Vinod Khosla

Business India, January 24-February 4, 2001

Ceasar: Vinod Khosla

Shivanand Kanavi

You can put my name in any search engine and you will get enough mate­rial on me, and I have said whatever I have to say in most of my interviews. So you can dis­pense with the usual questions and fire away," said Vinod Khosla, when we met him in his office at Sand Hill Road at Menlo Park. The words were not tinged with arro­gance but were a genuine attempt at getting to the core issues quickly.

That is how Vinod has made his famous picks: Juniper Networks, Cerent, Sierra, Redback and more. Which, according to Fortune, have made over $16 billion for KPCB, thereby making him "the most successful vc of all times". Clearly he has gotten to the core of the next generation of networking.

Vinod is famous for his brevity. Rajvir Singh, who has become a fountain of Optical start-ups, recalls how the first thing Vinod advised him in an e-mail when he invested in Fiberlane (later split into Cerent and Sierra) was: "Keep the B.S. out of all communication".

We say amen to that, and give below a few notes from our conversation, albeit pared with Occam's razor:

Money: In 10 years I have never done a rate of return cal­culation. I have only looked at economic contribution. After all, if you have made economic contribution, then money will come anyway. Many people talk about how much they will be worth. I reject all those who only talk about money. That is Wall Street mentality. It goes against my intellectual curiosity, predicting trends and so on.

Venture Capitalism: It is all about helping entrepreneurs build companies. Juniper is a classic example. When Pradeep Sindhu came to me, he had no business experi­ence. I guided him in building Internet routers and then helped him find the team, I helped him find Scott Kriens. All these things are really hard to do if you are just an engi­neer, because you have never done anything like this. What we do is help make an idea into a company. It is like being a coach for a soccer team or a football team.

Startups: I do not miss being in a startup myself. It is a lot of work and you get stuck in one area. Technology is mov­ing rapidly in so many areas and I have interest in so many areas. Every two to three years I completely change the area I am investing in. I take a few months off to learn the whole technology and develop a vision of what the world is going to be like - it is literally going back to school- then start investing.

Current interests:
Whether it is optical components, which is physics and material science or enterprise soft­ware, the only way to do it is to take three months off, learn and come back. My position lets me do it. I have got curiosity. I change my interests regularly when I get bored .All three of my degrees are in com­pletely different areas. Right now, as hobbies, I keep up with string theory and evolutionary biology.

Big vs small companies: It is not big vs small. People who refused to take risks are losing. Lucent had more talent than Nortel. But Nor­tel has changed: they have absorbed entrepreneurial culture. Lucent has wrong acquisition strategy and wrong culture. People don't leave Cisco when it acquires, but they do when Lucent does. It is much harder for big companies but Nortel has done it.

Optical Networking: In both opti­cal and wireless, valuations are hyped and over-hyped. But if you look at the impact they are going to have on society, on the way business is going to be run and so on, then they are underestimated. Investors are like lemmings, suddenly they go from greed to fear.

Indian entrepreneurs: The stockmarket is not a good indicator. Some have built businesses but some have built market caps. It is a bad value system. Issue is what can you create that has lasting value. Desh has real rev­enue. I like what Desh did. In the end his value will be judged if he makes an eco­nomic contribution. That is what Pradeep is doing. Intel, Sun, Dell, Microsft, Oracle all made contributions.

Education in India: A country of the size of India, a billion strong, does not have a major university which is world class and which is leading in research so that it does not have to depend on all the research in US. You have to take a SO-year view of this, not five to ten years. Over the long haul, India has the talent, language (English), enough infrastructure. It will grow in a very, very big way in the knowledge economy. Hopefully, people from all over the world will go to India to do research. That is the genesis of my interest in Global Institutes of Science and Technology.

Role models: I was 15-16 and living in Dehi Cantonment, as my father was in the army. I used to go to Shankar market and rent old issues of trade journals in electronics, which you get free there. I read about Intel being started up by a couple of engineers. That was my dream long before I went to lIT. In 1975, even before I finished lIT, I tried to start a company. Those days in India, it was not possible if your father did not have connections. That is why I resonate with role models. Andy Grove and Intel became role models for me.

Vinod Khosla loves travel and photography. Blown up pictures of his children taken by him are all over his office

Weather forecasting, Monsoon

The Weekend Observer, June 1992

Vagaries of weather forecasting

Shivanand Kanavi

The monsoon is not only a meteorological phenomenon for Indians but it deeply affects their literature, music, culture and the very psyche itself. Not only are the famous raga Megh Malhar and the poetic work Meghdoot expressions of it, but D.D. Kosambi, one of the great Indian encyclopaedists observed that the regularity of the cycle of seasons might have given rise to a fatalistic world view and even the myth of satyug, tretayug, dwaparyug and kalyug

Thus anybody who can predict and hopefully change the pattern of monsoons is always sought after, be they sooth sayers, astrologers, sadhus, performing yajnas and havans and even scientists. When the predictions or promised changes in weather do not come through, Indians seem to find any number of justifications for the failure of all the traditional wisdom but the meteorologist is never spared. He is the cartoonist’s delight. Who can forget R.K. Laxman’s cartoon of a long bus queue in pouring rain where everybody has provided himself with a raincoat or an umbrella except one fellow, and a guy whispering “must be from the weather bureau”!

All this is good for a laugh but when you meet a hardened weatherman like Dr. S Kumar, Deputy Director General of Meteorology heading the Colaba (Bombay) observatory and responsible for the western zone, and learn the rudiments of meteorology then you start appreciating the complexity of the subject.

The word monsoon owes its origin to the Arabic word mausam meaning season. It is believed to have been used by seamen, six or seven centuries ago to describe a system of alternating winds in the Arabian sea, these winds appear to blow from northeast for six months and from the southwest for another six months. Seasonal changes of wind are primarily the result of differences in the quantity of heat received from the sun by different parts of the earth.

As a consequence of its chemical composition and its soil structure, the conduction of heat into the earth is a comparatively slow process. Thus most of the solar energy received at the ground by the continents is used up in hating air rather than the earth’s surface. Whereas oceans are heated up to greater depths due to convection currents and a smaller part of the energy is available for heating the air, monsoon as a system of winds has the following notable features:
1. A system, with marked seasonal shifts, caused by the differential heating of the land and the sea.
2. A wind system that is largely confined to the tropics, that is the region between 20° N and 20° S latitudes on both sides of the equator.
3. Indian monsoon can be thought of as southeast trade winds which on crossing the equator are deflected to the right by the earth’s rotation (Coriolis force) and hence approach the land from a south-westerly direction.
4. The trade winds of the northern and southern hemispheres are divided by the Inter Tropical Front (ITF) which is a region of considerable cloudiness and rainfall. The southwest monsoon, originates in the ITF and moves northwards, due to the pull of a low pressure area in the hot Indo-Gangetic basin, but of the subcontinent after reaching the southern tip of India around June 1, it splits into two branches; the Arabian sea branch, and the Bay of Bengal branch.

The Arabian Sea branch gradually advances northwards to Bombay. The advance from Trivandrum to Bombay takes about ten days and is fairly rapid.

The Bay of Bengal branch moves northwards into the central Bay of Bengal and rapidly spreads over most of Assam by the first week of June. On reaching the Himalayan barrier, the bay branch of the monsoon is deflected westwards. As a consequence, its further progress is towards the Gangetic plains of India rather than towards Burma. The arrival of monsoon at Calcutta is lightly earlier than at Bombay. By mid-June, the Arabian Sea branch spreads over Saurashtra-Kutch and the central parts of the country. Thereafter the two branches tend to merge into a single current. The remaining parts of western UP, Haryana, Punjab and the eastern Rajasthan experience the first monsoon showers by the first of July. Some times the first showers at Delhi arrive from the east as an extension of the Bay of Bengal branch and sometimes from the south that is from the Arabian sea branch. Often it is a race between the two. By mid July it will extend to Kashmir and remaining parts of the country but only as a feeble current because by this time it has shed most of its moisture.

The normal duration of monsoon varies from two to four months. The withdrawal is much more gradual than its onset. Generally the monsoon withdraws from northwest India by the beginning of October and from the remaining parts of the country by the end of November. Though theoretically it seems possible for both the southwest monsoon and the northeast monsoons to co-exist in the southern half of the peninsula in October, in reality such situations are rare.

This in brief is the story of the monsoon but there are any number of disturbances of local and regional origin that can upset the text book schedule for example a cyclone in the Arabian sea that starts drawing the moisture away can lead to delays and dissipations.

The short term forecasts deal with a period of twenty four to seventy two hours which are mainly done with the help of data from over 500 weather stations spread all over the country, the data from the ships in the ocean, the satellite pictures from he NASA polar satellite NOVA which scans India every six hours, pictures from the geo-stationary INSAT satellite and even input from airline crews.

But a satellite picture, as Dr Kumar points out, is like a X-Ray photograph in the hands of a physician. It needs interpretation which is bound to be subjective. This is where the years of experience of our weathermen count.

Attempts are on to developing computer programmes to forecast weather in the medium term that is three to ten days at the Super Computer facility in Delhi.

The long term forecasting that is from ten days to a few months, is being attempted by the group in Pune. Over sixteen phenomenons all over the globe are being watched by this group and correlated with the Indian monsoon. Some of tem are the total snowfall over Eurasia during the previous winter, the convective wind between Darwin in the southern hemisphere and Tahiti islands in the pacific, the El Nino oceanic current off the coast of Peru in south America etc.

Considering the enormity of a weather system like the Indian monsoon, and the usual constraints of funds and technology and the very nature of a field where controlled experiments are well nigh impossible, our weathermen are doing a competent job, to say the least.