CHARLOTTESVILLE, Va., April 14 -- The optical access market will grow from $2.1 billion in 2003 to just over $3 billion in 2007, driven mainly by growth in Ethernet, course wavelength division multiplexing (CWDM), passive optical networks and free-space optics, according to a study, "Access 2003: A Market and Technology Assessment of the Broadband First Mile, " by telecommunications industry analyst CIR. CIR said simply giving startups hundreds of millions does nothing to reduce costs, and the act of funding new technology by itself is not the answer, whether it's wireless, passive, or robotic. "When considering how cheap old technology like Ethernet has become, it is not the result of design or listening carefully to customer requirements," CIR said. "It is a product of time and patience. The 30-plus year history of that technology has allowed the creation of tremendous production volumes, which are the leading cause for Ethernet's low unit costs. Patience will allow time for the opportunity to generate volume and, in turn, to reduce unit costs for free-space optics, 802.11 and other promising technologies. New requirements, like security, can then be layered above without destroying the volume efficiencies." CIR said no single startup or unique technology will provide 1999-style returns. "However, one issue is clear: no service provider, large or small, can match the production scale and technology cost reductions a vendor achieves from selling to thousands of corporations. Ethernet and CWDM are examples of this, and also the reason why free-space optics startup Terabeam is now selling equipment directly to business customers." CIR said if AT&T implements free-space optics or some other access technology, "there is certain to be a flood of press releases, news stories and analyst quotes. However, it is the decisions of CTOs and IT managers at thousands of universities, hospitals and brokerages that are the real determining factor for any new technology's success." For more information, visit: www.cir-inc.com