The EISA announcement could produce one of several outcomes. We think the most likely outcome is that EISA and MCA will coexist, with neither architecture able to deliver a knockout blow on the other. However, there are other possibilities. None of them has disastrous implications for Apple.
Scenario 1. EISA wins big. EISA could be embraced by most PC customers, and become the de facto standard bus for 32-bit Intel-based PCs. If this happened, IBM would face two alternatives:
• Stick with MicroChannel, and be content to supply PCs primarily to a set of captive customers (perhaps as much as a quarter of the US PC market—a lot in absolute terms, but less than IBM’s current share). This might allow IBM to charge higher prices for its PCs if it improved MicroChannel quickly, and would let it control the accounts that did choose to buy from IBM.
• Join the EISA group and produce its own PCs using that bus. This would leave IBM in the same position that it faced with the old PC architecture: little proprietary content to justify higher prices and ensure account control.
If the EISA group could start manufacturing its new bus immediately, we might give this scenario a good chance of coming about. However, the expected 12-month delay before the first EISA machines ship will make it very tough for them to dominate the market. IBM will have a huge installed base of Micro Channel machines by the time the first EISA PC is sold.
Ironically, if EISA succeeds wildly, it will probably have little impact on Apple. Our own plans have assumed that there would eventually be a single standard 32-bit bus for PCs; if it’s EISA instea of MCA, that makes little difference to us.
Scenario 2. EISA and Micro Channel coexist. The EISA manufacturers and IBM are evenly-matched in many ways. The EISA group has a larger total share of the PC market, while IBM has a big installed base of MCA machines, and the magic of its name. This could result in both EISA and MCA machines sharing the market. Shares might remain about where they are today.
The big losers in this situation would be customers. They would have one more confusing decision to make, and would have to deal with two bus architectures if they mixed EISA and MCA machines. Incompatible architecture is one of the main reasons corporate customers give for not buying Macintosh, and we think they will be little more tolerant of that sort of problems with Intel-based PCs. Therefore, we think most companies will probably standardize on one architecture or the other.
The main impact this situation could have on Apple would be if customers viewed their purchasing choice as EISA vs. MCA, instead of Macintosh vs. PC-standard. If the decision becomes too complicated, some customers mught choose to simplify it by eliminating Macintosh from the picture altogether. For this reason, Apple may want to emphasize the similarities, rather than the differences, between MCA and EISA.
Scenario 3. Micro Channel wins big. In opposing Micro Channel, Compaq and other vendors have repeatedly argued that customers just don’t need a 32-bit bus at the moment. Now that Compaq has endorsed a 32-bit bus, customers might decide to buy the only one available today: MCA. We think IBM is likely to try to encourage this train of thought. If customers go along enthusiastically, Micro Channel sales could actually rise. However, we think this is unlikely; customers who trusted Compaq yesterday will probably continue to trust them.
Scenario 4. IBM licenses MicroChannel openly. At least one major EISA vendor has hinted privately that it would abandon the consortium if IBM agreed to license Micro Channel at a lower costs. If IBM did so quickly, it might be able to kill EISA. However, IBM’s window of opportunity is small; the longer that EISA exists, the more prestige and publicity effort the members will have invested in it. They will eventually have too much of a stake in the consortium to abandon it, no matter what IBM does.
Licensing Micro Channel freely would be almost as bad for IBM as adopting EISA. Although IBM would control the standard, any innovations it made would belong to every other licensee. IBM would be unable to differentiate its products on the basis of bus architecture, and would once again be subject to severe price competition.
From Apple’s point of view, the main effect of widespread Micro Channel licensing would be a faster diffusion of 32-bit buses into the Intel-based PC world, since Micro Channel is already a finished standard. This might speed up slightly the transition to full 32-bit operating systems, and could reduce customers’ confusion and make them feel more comfortable about buying PCs. However, the direct impact on Apple would be slight.
Scenario 5. Customers stop buying. OS/2 and Presentation Manager software will take years to arrive in fully functional form. Now a large number of PC makers will be arguing in the press that customers should wait a year for EISA instead of buying MCA today. Given the level of cunfusion that will be produced by this situation, we think many customers may decide to scale back PC purchasing and wait for the PC-makers to get their act together. If this were coupled with a mild business downturn, it could produce a dramatic flattening in PC sales.
Scenario 6. Customers turn to Macintosh. Apple already has a 32-bit bus architecture, shipping today. Apple already has a windowing operating system, and a large pool of software that works with it. If Apple can communicate to customers the message that it is selling today the sort of solutions the PC industry will have in a few years, we may be able to attract new customers to Macintosh. To the extent that EISA increases the level of confusion in the PC world, it makes Macintosh seem a more rational, better-organized alternative.
