Exent: Bringing games to the Internet

Exent Technologies deals in Internet streaming applications, i.e. real time communications that can be conducted without downloading computer files. This field is one of the most critical for communications companies. In order to make money from the Internet, you have to provide applications and services that subscribers will pay for. The most common applications are video, games, and music.

”Exent actually helps create a market,” says founder and CEO Zvi Levgoren. “Luckily for us, the basic demand already exists. In contrast to many Internet services, which tried to create a demand for a new service, we’re taking an existing demand and simply trying to streamline it and provide a much better solution.”

The surprising aspect of Exent is its claim that the most popular application at the moment is computer games, not video or music. One figure supporting this is that 78% of all CD-ROMs sold worldwide are games. 74 million games were sold last year; 12 million households bought an average of six games each. This industry was worth $63 billion in the US alone. Levgoren says that, according to his data, “In the US, this industry is bigger than movies or music.” Exent is attempting to gain a foothold in the market.

Internet-based computer games are actually still a futuristic vision. High-speed Internet is only the beginning of the technological requirements, since computer games combine all 10 multimedia technologies, including sound, images, video, and interactive communications, according to the number of users. It was once thought that the Internet games market would mature in 2002, but it now looks like we must wait a longer time for these games to be available commercially. Levgoren wants to be prepared, with leading technology, which is why he is building Exent.

Levgoren knows the importance of lying in wait for demand. He spent most of his years in the high tech industry in a company that every reader of the economic press now recognizes as one of the three biggest Israeli successes. Comverse (Nasdaq: CMVT) made its big breakthrough by betting on cellular communications, the winning horse. It waited for the market; and its management had the daring and the brains to take advantage of the connection between its technology and the market, allowing it to become a major company.

Betting on the computer’s brain

In 1992, Levgoren set out on a new path. The venture capital industry was an unknown concept, both in the press and in reality. Another source of capital was the Chief Scientist of the Israeli Ministry od Industry and Trade, but Levgoren was not interested: “Support from the Chief Scientist is a negative factor. It delayed Comverse’s development. The Chief Scientist is a source of inefficiency in a system. I believe, however, that Carmel (Chief Scientist Carmel Vernia – M.B.) is aware of this and will change the situation. We therefore raised money through the grocery store model. We sold products and invested the profits in the company.”

The company Levgoren founded was called ARMT Multimedia and Telecommunications. The contrast between Exent and the Israeli-sounding ARMT indicates the cultural-business change since those days, when financial angels were already around. With the help of such an angel, David Katzir, who financed the company for its first ten months until its first order, the company survived. Levgoren: “We dealt mostly in the digital video field – how to record video on a disk and manipulate it. This field began to develop at the time with companies like Optibase (Nasdaq: OBAS). We dealt more in manipulations, rather than processing specific information. Two to three years later, this field led us to high-speed communications.”

”If we ignore the efforts of VDOnet, the video field led in the direction of high-speed broadband communications. We focused on transmitting heavy files through a broadband pipeline. Our customers at the time included the Ministry of Defense, all the Israeli banks (except for Bank Hapoalim), and Bezeq, which was starting to study high-speed communications when we focused our development on the application level – what to do when you already have a broad pipeline. I first heard the term ‘high-speed Internet’ at the 1996 CeBIT exhibition.”

Then as now, however, the key question is which applications are most in demand – narrow band or broadband. As the leader of a company that sought from the beginning to deal in applications, rather than with the installation of the pipeline, Levgoren struggled mightily with this question.

”We always thought that vertical applications would lead in broadband,” he says, “so we tried to develop online music with digital CD quality. Some of this activity was a feasibility study – (determining) which applications were in demand and which weren’t.” Another lesson learned by ARMAT at the time was that different applications on the same server require varying amounts of bandwidth - even if they use the same number of bits - and therefore reach the customer at different speeds.

This fact is of enormous importance in determining the ways that service providers plan and construct a network from end to end. Levgoren explains, “This fact actually has great consequences for all aspects of communications network construction. I think that every home market communications provider should examine what the five most popular applications are – the ones for which the consumers will agree to pay a lot of money. The network should be designed for optimal service of these specific applications.”

In the course of examining various applications, ARMT came to the field of games: “The beginning of (our involvement with) the games matter occurred during a project we did for Bezeq, examining this application. The technological challenge initially appeared impractical – transmitting CD-ROM quality games over a broadband communications network.”

”Globes”: What was the problem?

Levgoren: ”As always, reality. The entire broadband concept proved to be much narrower band and slower than we anticipated. In that prehistoric period, we thought that a 2 Mbps bandwidth was realistic. In practice, we saw it wasn’t and could never be practical. The question was how to transmit games, given these figures. The first solution was compression. That’s the type of solution used by standards like JPEG and MPEG. But what should be compressed? The graphics of the game itself are already compressed, while the uncompressed vocal component of the game is insignificant. If we compress the voice and lose some bit somewhere, the enjoyment is gone and the game is wrecked.”

So what is the solution?

”We decided to examine the entire problem from a different point of view. Instead of trying to optimize the reading of the information from the disk, we examined whether the CPU (Central Processing Unit), the brain of the computer, reads the data in the same way.”

And you discovered it was not the same.

”Exactly. We found out the CPU requires different quantities of information during the game, not a constant speed. At peak points, a maximum speed of 10 Mbps is needed, while there are practically no requirements at other times.”

From this theoretical understanding you moved on to a technological application?

”We understood that if we manage to build a format for a given application, we could overcome the problem by sophisticated planning of information transmission and various manipulations. This would enable us to handle peak bandwidth requirements by taking advantage of moments of low demand. The trick is to find a profile by which the CPU can read a given game and use manipulations to transmit more information while it’s reading relatively little. For example, in some cases a game begins with an introductory video film without a wave format that has peaks and lows. Our manipulation would be to show the film minutes before use of the game begins, while performing slight manipulations of the content, such as reducing the upper and lower margins of the introduction in order to reduce the bandwidth required.”

So your technology applies only to applications with a wave format?

”Precisely. Our technology is designed to handle applications with a wave profile – those that must be handled by the user, the software that operates the brain of the computer.”

End of the grocery store model

Up to this point, ARMT operated under a grocery store model. They sold, made a profit, and invested it in continued development. They now realized, however, that their technology had much greater market potential than they had previously harnessed. It was time to take advantage of the spirit of the times – the new business models reaching Israel from Silicon Valley. In short, Levgoren and his pals understood they had to raise external capital from venture capital funds. They took the opportunity to change the company name to a younger, more American one, which would refer to the transmission of exe files on the net – Exent. Levgoren: “In August 1999, after seven years without raising capital, we raised $5 million: $4 million from Concord and $1 million from Bezeq, which converted its former rights into shares and currently holds 5% of Exent.”

”The next financing round took place when Concord heartily recommended raising money in the US. As with Yair Safrai’s involvement as a director on our board, they didn’t confine themselves to theory. They helped us get investments from the NEA funds (New Enterprise Associates – a $2.3 billion fund considered the strongest on the US east coast – M.B.) and General Atlantic Partners. There was actually a kind of tender between the two funds and we preferred NEA, but General Atlantic insisted on investing. They did it through the fund of the management company’s partners, which made the investment in us still more important for them. Our main strategic investor is America Online (AOL), together with Comcast Interactive Capital (CIC), the third largest US cable company. We get help from AOL on a daily basis.”

What help do you get from NEA, which has quite a large investment portfolio?

”Since we come from a relatively small country, we’re incapable of grasping the extent of NEA’s networking. They help us from day-to-day. With their assistance, we can reach places I couldn’t even dream of reaching with normal marketing.”

Who are your competitors?

”We have three competitors: two current ones and one potential competitor. The most dominant is a private US company called Into Networks, a larger company than Exent, with much stronger positioning in the cable field. We’re going up against them head to head, but I can’t recall ever losing to them.”

”Our second direct competitor is Infogate, an Israeli company in which Bezeq also has a stake, as well as the Zisapels. I can say we’re not afraid of them. As far as we’re concerned, they’re simply not there in the market. They’re not reaching the final stages of trials, because their technology level is just not good enough.”

”Our potential competitor for the future is AppStream. They deal in streaming applications, which puts them in a good position for becoming our competitor.”

Published by Israel's Business Arena on June 14, 2001