October 15, 1995 Dear Kirk, Thank you for the opportunity to be part of this exciting project. Per your request I have evaluated the options for enhancing the production capacity to maintain and further develop our market share. The semiconductor market in business communication is expected to have a compound average growth rate of almost 18% over the next five years. Our current market share within this segment is slightly more than 7%, and management wants to see the share of the market double by the end of the century.
Figure 1 represents the projected growth of the semiconductor market in business communication and Mitel Semiconductor over the next five years. [pic] In order to obtain the management goal, we need to expand our production capacity. Four options are available to increase capacity, and they will be evaluated by the following factors: ? Cost of investment, ? Market demand, ? Fitness to the company’s immediate and long-term requirements, ? Risks involved, ? Overall operational benefits in each option. Option 1: Status Quo
With this option, we would continue the production of 100mm wafers, but would increase the wafer capacity by 44,800 and decrease the line width from 1. 2 micron to 0. 8 micron. Positive: • Upgrading would cost $10 million; relatively low compared to the other options. • The upgrade would be accomplished in eight months without a plant shutdown. • 100mm equipment could be purchased for low cost from other FABs switching to larger wafers, and we could train our own technicians for maintenance. • Profitability might be sustained for the near future by maintaining low production costs, provided we can obtain a supply of wafers.
Negative: • Wafer supplier indicated that they are going to convert to a larger wafer size. As a result of the reduced supply of 100mm wafers, we would have to develop in-house wafer-growing capability for an additional cost of $40-50 million. • Most equipment manufacturers no longer produce 100mm equipment, so obtaining spare parts and servicing existing equipment would be more difficult. This option would not increase the current capacity substantially. We would be using the 100mm wafer size, and the wafer would have to be produced in-house for an additional investment.
The total cost of investment would be $50-60 million. This option would require us to make another expensive conversion at some future date to meet the demand. Option 2: Convert Bromont to Larger Wafers This option would require production equipment changes to produce larger wafers. The line width would be decreased from 1. 2 micron to 0. 8 micron. The larger wafer size options would be: 150mm, 200mm, and 300mm. a) Conversion from 100mm to 150mm Positive: • The capacity would be five times the current capacity. • Total cost of investment would be $35-40 million; relatively low compared to the other options. Relatively short conversion time would be needed compared to 200mm and 300mm; would take two years without a plant shutdown. Negative: • Although supply is plentiful and cheap at the time, it would not take long time for wafer manufacturers to stop producing 150mm wafers. • Another expensive conversion would be needed at some future date. b)Conversion from 100mm to 200mm Positive: • The capacity would be nine times the current capacity. • The useful life of the manufacturing plant would extend. Negative: • The total cost of the investment would be $150 million. Current equipment would not be used, and there is a limited supply of FABs. • Four years conversion time would be required. c) Conversion from 100mm to 300mm Positive: • The capacity would increase substantially and it would be 20. 25 times the current capacity Negative: • The total cost of the investment would be $250 million • The technology is immature [pic] Figure 2 represents the investment/capacity increase in conversion to larger wafers. The best option in terms of investment/capacity increase would be 150mm in conversion to larger wafers. Option 3: Contract Out Fabrication
With this option, we would increase our capacity dealing with a supplier. Positive: • The capacity would increase with no initial investment. • The capacity would be the same as demand. Negative: • Fixed costs would increase by $1 million annually. • Cost would increase to $600/wafer, which would cut our gross profit by 35%. The cost calculations are given in Exhibit 1. • There would be a high risk to secure the capacity, which would require large up-front payments. Option 4: Secure Capacity Though Acquisition This option would include the acquisition of a small FAB in Jarfalla, Sweden wned by the large engineering firm, ABB. The current facility is for R&D, but could be integrated into Mitel Semiconductor line. Positive: • The cost of investment would be $45-50 million. • Security of supply for customers would be provided. • Could take advantage of the current market of the FAB, and create more demand. Negative: • Capacity would not increase significantly; it would increase by 20,000 immediately, and could be brought up to 48,000 in twelve months • The culture would be different since this facility is for R&D. Acquisition has a high failure rate. • There would not be any capacity relief for Bromont until the production would be increased. • We would still produce 100mm wafers, so there would be a reduced supply of wafer, and additional investments would be needed Evaluation of Options Exhibit 2 shows the future capacity of each option. After taking into account the above analysis, the following are the most significant: • Exhibit 3 shows that option 1 and option 4 would achieve a capacity under the demand, so they would not be long-term investments. While option 2 and option 3 would surpass the demand line, option 2b and option 2c would reach a significantly higher demand and would have a much greater cost. It would not worth to make great investments to the newest technology in a market in which the technology improves rapidly. • Option 3 would be too risky because of the lack of security for capacity. It would also increase the cost to $600, which would cut down our gross profit significantly. The Decision The best option to suit our needs is option 2a: to increase the wafer size to 150mm.
Conversion to 150mm wafer would meet the demand, and would require lower investment than the other options. Since this option would take 2 years to be accomplished, we would need another immediate action to meet the demand till the accomplishment. Immediate actions could be option 3 and option 4, since they can increase the capacity immediately. Option 4 would not relieve the capacity for Bromont for a year until the production would be increased, and there has to be $35-40 million investment. Option 3 would require only a fixed cost of $1 million per year, and the cost per wafer would increase to $600.
As a result, I have concluded that option 3 would be more advantageous for the immediate response, in which we would contract out the fabrication for the surplus demand, while we take the necessary steps to insure we capitalize on the growing demand in the semiconductor market by increasing our production capacity in the long-term by conversion to 150mm wafers. Exhibit 1 |Mitel Corporation Product Revenues |$521. 7M | |Mitel Corporation Products Cost of Sales |$275. M | |Mitel Corporation Cost/Revenue |53% | | | | |Mitel Semiconductor 1995 Sales* |90,000 | |Mitel Semiconductor 1995 Revenues |$77,700,000 | |Mitel Semiconductor Price/Wafer |$863/wafer | |Mitel Semiconductor Cost/Wafer** |$456/wafer | |Mitel Semiconductor Gross Profit |$407/wafer | |Mitel Semiconductor Gross Profit with $600 supply |$263/wafer | |Decrease in Gross Profit |35% | *Assuming that the given 1994/1995 sales is for 1995 sales **Assuming that Mitel Semiconductor has the same cost/revenue ratio as Mitel Corporation Exhibit 2 Option # |Capacity |Option | |Option 1 |352,800 |Status Quo | |Option 2a |567,000 |100->150 | |Option 2b |1,008,000 |100->200 | |Option 2c |2,268,000 |100->300 | |Option 3 |414,700** |Contract Out | |Option 4 |160,000 |Acquisition | *Capacity calculations are calculated as: • The number of chips producible on a wafer varies as the square of the wafer diameter • The number of chips available from a wafer goes as the inverse square of the line size **In option 2, it was assumed that the decrease in line width from 1. 2 micron to 0. 8 micron would not require any additional cost. ***Capacity of option 3 is equal to the demand at 2000. The demand at 2000 was calculated as: [pic]. It was multiplied by 2, because management wants to see the market share doubled by 2000. Exhibit 3 [pic]