Apple Turns Macintosh After IBM PC AT
The Journal Record
October 2, 1985
Apple Computers, Inc. is lining up suppliers for new computers and expansion options, the bulk of which are expected to be introduced in January.
Apple has contracted with Rodime to supply 100,000, 20 megabyte Winchester disk drives for the Macintosh. Apple has also signed an agreement with Sony to supply it with 250,000 800K double sided 3 1/2 inch floppy disk drives before the end of 1986.
For several years Apple has been known to be working on its own Winchester technology and the manufacturing of its own hard disk drives. This project has been discontinued and Apple is now relying on third party suppliers for its disk technology.
Industry analysts suggest that Apple is preparing to produce a version of the Macintosh aimed squarely at the IBM PC AT market - one that would offer as significant a differential in power over the AT as the Macintosh did to the PC. With applications like Jazz now available for the Macintosh, the desire for a hard disk and more power is common among users.
Currently there are several options available for those who desire a hard disk but they are radically different from the options presented to an IBM PC user. Today, a 20 megabyte hard disk that plugs into the IBM PC is available for as little as $495. It can, in most cases, be installed by the average user.
The Macintosh, however, provides a serial port interface for a hard disk, which is fine except for the speed restrictions of a serial port. In most instances, the serial port on the Macintosh would cut the performance of the hard disk drive by one quarter to one half. It makes things like auto booting off the hard disk (customary in the IBM PC world) very difficult.
One company has discovered a way to mount the hard disk drive directly in the Macintosh case and avoid the interface problems.
General Computer has announced a 20 megabyte Hyperdrive designed to fit internally in the Macintosh. This is accomplished via the desoidering of the 68000 and the installation of some customized interface circuitry.
General Computer Corp. has had to make special arrangements with Apple just for the installation of the hard disk drive in order not to void the warranty. They have also had to develop special software that divides the hard disk into 32 file drawers so that the Macintosh's operating system will not be overwhelmed with files.
But in my mind, worst of all is the price which at $2,795 is five times that of the least costly method for an IBM PC.
- How powerful will the Mac II be? One of the best indications for the power of a personal computer is the microprocessor on which it is based. The Macintosh II is expected to be based on Motorola 68020 that provides an upward code compatibility with the microprocessor the Macintosh is based on.
When it comes to power, the 68020 is at least five times faster than an IBM PC AT. In other words, those who assert you can service 16 users on an IBM PC AT would lay claim to 92 users on a Macintosh II.
While this will not be the case, technical specifications released by Motorola suggest that it should very easily be able to support 16 users and provide each of them with approximately the processing capability of an IBM PC. If Apple chooses to base its Macintosh II on the latest in silicon technology, then they will either have to develop their own multitasking multiuser operating system or use Unix.
In the dark ages of computer science (three or four years ago), a standard industry price performance ratio was set at around $100,000 per MIPS (millions of instructions per second). The 68020 provides a 2.5 MIPS or 8.3 MIPS burst rate, meaning that three years ago many would have argued that the new Macintosh should be priced at $250,000. I anticipate it will be priced at less than 1/50th of that.
The Macintosh II is also expected to provide the option of a floating point coprocessor. Such coprocessors can dramatically speed up the performance of things like spreadsheets and calculation oriented applications. The AT provides such an option with the 80287. The numeric coprocessors in the Macintosh II will run rings around the coprocessor in the AT, not to mention the less powerful one in the IBM PC.
Comparing arithmetic done with the floating point processor in the Macintosh II to an unaided IBM PC without a numeric processor, a 10,000 percent increase in the speed of some applications is a reality.
It has been said that writing in assembly language on the 68020 resembles basic more than does the assembly languages of the past. The 68020 is also capable of supporting up to four gigabytes of memory, which is 4,000 times greater than the maximum memory configuration of an 8088 based machine like the IBM PC, an equivalent to two million double-spaced typed pages.
At the risk of making a foolish statement akin to that made at the introduction of the IBM PC - that 640K (320 pages) would be enough memory for any application conceivable at that time - I suggest that four gigabytes is enough memory for about 100 of the largest applications I can imagine being written to run simultaneously, something this microprocessor, although powerful, would not be able to support effectively.
Apple will undoubtedly improve the display technology in the Mac II. They basically have two choices. They can increase the resolution of their monochrome screen and make it six times as large by going to something with a 1024 by 1024 display or go to a color system like the Amiga. My guess is they will go to the ultra high resolution monochrome screen while providing for an Amiga-like color option.
Apple must also figure out a more efficient way to interface its peripherals than through its serial port. This calls for the tried and true buss architecture of the Apple II, IBM PC and AT.
Apple has made great strides in user friendliness and, with such a powerful machine, applications that learn how the user works would be possible. For instance, if every week you loaded your word processor and did some editing and then sent out your work over MCI mail, the computer might eventually figure this out and do as much of it as it can on its own. It is also easy to imagine an operating system that would correct typing mistakes as they happen.
For under $5,000 the Macintosh II will be a machine as powerful as a VAX and sit on your desk able to run circles around your AT.
A machine like this can be built today and, with the economies of scale expected to be reached in the next six months, be delivered for under $5,000. In fact, it will eventually be marketed for much less.
The lacking feature might be software, but since the 68020 offers compatibility with the microprocessor the MacIntosh is based on, for single user applications this should not be a problem.
Developing a multiuser multitasking operating system can be quite a task in itself. If Apple decides to develop its own proprietary system, it could take a year or two extra. I question whether Apple desires to build such a machine. That is unclear today.
The Macintosh represents the vision of one man, Steve Jobs, who for the most part can be credited for both its good points and its bad. Most notably, Jobs does not believe in expansion slots to the extent that they are found in the Apple II and IBM PC computers.
Although a cofounder of Apple, he has now been ousted by a man who would like to see a Macintosh more like the IBM PC. Jobs has just raided Apple for some of its best engineering and managerial talent in order to start a new company that would be funded largely by the proceeds from the sale of his Apple stock. The new company reportedly will develop powerful work stations/personal computers for the educational market that may be much like the one described above.
This has caused a fury within Apple and led to Jobs resigning as chairman of the board.