The AVALON – graphics chip is the optimal translation of software to hardware.

Nobody today can produce a chip without first having a clear and precise picture of which algorithms the chip should contain and how these must be implemented.

The algorithms which AVALON will incorporate are to a large degree already implemented in sofware. Using this software the algorithms may be tested and each sub result relatively simply confirmed.

By using this software a chip design can be produced which essentially executes the same calculations as the software. Bearing in mind that a logic circuit is only marginally comparable to programming, substantial optimisation is required to keep the complexity of the chip to a minimum (cost), achieve low power consumption (later deployment) and above all achieve the required performance.

Since generally only one objective among many can be optimally achieved, the focus will be placed on performance. The performance must be exorbitantly high and can only be attained by special parallisation and detailed knowledge of the algorithms employed. Approximately 300 to 500 parallel calculation and storage operations must occur. The coordination of these operations alone demands a high level of expertise. Development support is being provided by the University of Aachen.

It is planned to start the draft of the AVALON chip this year (subject to having identified an investor). Three to four experienced hardware designers will participate at this stage.

The draft will be completed by end 2003 and transferred to the service department of a chip manufacturer, who in turn will require approximately 6-9 months to prototype the first AVALON. During this time various interfaces to operating systems to support AVALON will also be completed. Additionally, certain chip manufacturer requirements could emerge which need to be rapidly integrated in order not to jepardise the plannned production date of early in 2004.

Once the AVALON prototype is completed and functional in early 2004, first prototypes must be sent to graphics board manufacturers for testing. The selection of the graphics board manufacturer can be determined in advance by the mid of 2003.

The market situation amoung graphics board manufacturers is highly tense due to extreme competition and the need to rapidly innovate due to the lack of differentiators. Consequently, it will be important to identify and convince a partner company to create a lead into the market. The goal is to find, as a first step, two graphics board manufactures willing to implement AVALON. If everything goes to schedule, the first graphics boards with AVALON can be on the market by mid 2004.

A market share of ca. 0,5% is targeted within the first two years. This translates into one million of the 200 million graphics chips sold world wide coming from RaytracingTechnology. By aiming for a unit price of $100 for batches of 10,000 AVALON can achieve margins of $10 to $20 per chip. Total development cost for the fist generation chips, approximately $15 million with chip manufacturer costs accounting for $600,000 to $1 million of this cost, would then be covered.

Sales are expected to grow after the introduction year as the demand for graphic features in future applications increases with AVALON providing one of the first generation products in the market. It is expected that a market share of 1% can be substantially exceeded.

The development plan envisages starting design and implementation of the next generation chip from mid 2004 in order to maintain the market lead. RaytracingTechnology will need to expand operations to meet increasing demand for customer support and to maintain software quality. Latter can be regarded as a strategically important factor for business development in view of the expierences of many users with average quality and performance graphics products today.

All of this will not be simple, but were it simple it could be produced by many.......