Yesterday, “Globes” reported that EZchip Technologies existing investors injected another $13.5 million in the company, and will inject a further $8 million later. A $3 million credit line has also been guaranteed.
On one hand, one could claim that the financing will bring the company to profitability. On the other hand, general disillusionment with the company’s sector, network processors, has caused many companies to either abandon it or at least suspend product development. These include start-ups like Entridia, which sold its remains to Stratigos Networks, which had already quit the field; TeraGo Networks of Canada, one of the first companies to operate in EZchip’s 10 gigabyte field; and Acorn Networks of Japan. Established companies like PMC Sierra (Nasdaq:PMCS) and Vittesse Semiconductor Corporation (Nasdaq:VTSS) have announced they are quitting the field.
EZchip develops network processors that are placed inside routers or other network communications equipment. The technology’s main advantage over ASIC-designated chips is its flexibility. If customer wants to make changes in a chip after its development is completed, it turns out to be impossible. Network processors, in contrast, got investors excited because they can be adapted for specific tasks. In addition, they are relatively cheap and can be programmed after the fact. This promise was translated into $500 million investments in start-ups in the field.
The five largest companies in the field are Motorola (NYSE:MOT), Agere Systems (Nasdaq:AGR) [a Lucent Technologies (NYSE:LU) spin-off] IBM (NYSE:IBM,( Intel (Nasdaq:INTC) and Applied Micro Circuits Corporation (Nasdaq:AMCC). One of the largest-ever acquisitions of Israeli technology was related to this field, when Applied Micro Circuits acquired MMC Networks for $4.5 billion in shares.
But that was way back in 2000. Since then, many bytes have flowed through the Internet, and forecasts of $1 billion sales a year for network processors have since dwindled to $150-200 million, shared among all suppliers, with the four leading corporations dominating the market. An increasing number of companies have decided to rely on ASIC chips, some claiming they can even compromise and use ordinary processors. Under these circumstances, companies and especially investors stop to think and decide how much longer they can wait.
What plays in EZchip favor in this unstable climate is its technology, which is still ahead of most of its competitors, even the largest among them. But this fact is not enough to guarantee a rosy future. The company’s only officially declared customer is China’s ZTE, although EZchip claims it has 13 more customers generating several hundred thousand dollars in revenue a year. The market is still too small to depend on, which explains the latest financing round in which no new investors participated.
As for acquisition, a $4.5 billion price tag in a $200 million market is not a probable scenario, even if one could imagine that one of the major companies wanted to buy EZchip to gain a technological advantage. The leading candidate is IBM, which already owns 5% of EZchip, and whose plants make its chips. On the other hand, this relationship will not necessarily lead to preference as a buyer. IBM also invested in optical communications company Kymata, which was recently sold to Alcatel (NYSE:ALA).
Could the next acquisition come from the other direction entirely, with EZchip the buyer? Our bet is that EZchip will reinforce its technology by buying search engine technology. This doesn't refer to Internet search engines, but to software integrated into semiconductors that quickly locates information streaming routes. The immense amount of information handled by these chips requires this type of traffic control.
So far as is known, EZchip is currently using an external search engine, but integrating the technology in the chip would give it an edge making the company much more attractive. There have been two recent interesting acquisitions in this field: IDT (Integrated Device Technology, Inc.) (Nasdaq:IDTI), acquired Solidum Systems (in which Vertex Management Israel an investor); and Cypress Semiconductor (NYSE:LYR) acquired Sahasra long after Cypress quit the network processing field.
If EZchip uses its capital to make such an acquisition, in our opinion the leading candidates are Israeli start-up HyWire or FastChip of the US, since BestRoute will probably be acquired by IBM or Agere first.
EZchip’s next chip will manage traffic. This is even more of a niche chip, designed to examine the ratio between an incoming data packet and its objective in the supplier’s service package menu where the chip is installed. In other words, incoming data packets that belong to the service, such as VoIP or video-on-demand, jump to the head of the queue. It will also require a lot of public relations to explain why such chips are necessary, compared with ASIC chips, especially since they do not replace network processors, but sit alongside them. Israeli start-up Dune Networks, which is active in this field, made its plans known a few months ago, and has $24 million in backing.
Competition in the field is already intense. Applied Micro Circuits is selling traffic managers, Agere will launch its version in the first quarter of 2003, California-based start-up Internet Machines has developed a combined network processor/traffic manager, Motorola is still stuck at the 2.5 gigabyte stage, claiming that is sufficient, and there are plenty of other start-ups in the running.
Published by Globes [online] - www.globes.co.il - on December 18, 2002