Worth switching to the new microcomputers?
Should microcomputer users stay with their current units, upgrade them, or switch to the new models? The decision is important because of potential incompatibility in software, hardware, and data between different types of computers.
Let's look at the IBMs first:
* The old IBM PCs. The PC and the XT (like a PC, with a hard disk) use the Intel microprocessor 8088 at a speed of 4.7 megahertz. The AT, using the Intel 80286 microprocessor at speeds of 6 to 8 MHz, is about two to three times faster than a PC or XT. In addition, it uses floppy disks with a capacity of 1.2 Mb (million bytes), versus 360 Kb (thousand bytes) for floppies on the PC and XT. The higher capacity makes it easier to back up a hard disk and maintain a library of copies. The AT also can access more memory and run larger programs.
A number of companies imitatethese IBM models, and they do it so well that the imitations are called clones. Many clones are excellent, some are not too good, in the experience of computer magazine reviewers and users. The clones usually run two to four times faster than the original, and you often can buy two of them for the price of a single IBM.
* The new IBM PS/2. The Personal System /2, which comes in several models, is IBM's answer to the technically superior clones. Model 30 (PC/XT replacement) uses the 8086 microprocessor and is 21/2 times faster than a PC or XT. Models 50 and 60 (AT replacement) use the same 80286 microprocessor as the AT, but they run twice as fast. Model 80, with the 80386 microprocessor, is 3 1/2 times faster than the AT and has the potential to operate many of its programs almost simultaneously.
In addition to their faster microprocessors, all of the PS/2 models include a new 3.5-inch floppy disk format with a capacity of 1,440 Kb, more integrated circuits, and a different circuitry (called Micro Channel) to interface accessories. With the Micro Channel structure, data can move faster between different parts of the computer. A disadvantage is that the circuitry will not work with boards available for interfacing with the IBM PC; new boards will have to be created.
Some of the more advanced features of the new operating systems, including simultaneous running of tasks, will not be available for quite a while. Although some current software may not work with the new operating systems, programs developed for the PS/2 models will be more powerful offer more possibilities-because the computers are faster and can access more memory. The computers will also provide better graphics.
Clones of the PC and XT run as fast as PS/2 Model 30 (the PC/XT replacement) for half the money, and clones of 3.5-inch drives are already on the market. Clones of the AT run as fast as or faster than the Models 50 and 60. A further advantage of the clones is that they are compatible with the original IBMs.
Because IBM delayed its introduction of the 80386-based PS/2, many other companies have been able to reach the marketplace at the same time with their 80386based microcomputers. Almost all of these models are compatible with the AT but not with 80386-based PS/2. Clones of the latter are likely to appear in the near future.
My conclusion is that the PS/2 may, be a system for the future (one to four years), but if you need something now it offers little, if any, advantage over the old IBM microcomputers and their clones. If an application requires a speed and performance (higher memory, different drives, etc.) that cannot be provided by the old IBMs, you can choose between a clone or a PS/2. The PS/2 will cost you about twice as much, is not compatible with existing boards, may not take some software, and may or may not work for your existing applications.
Prices on the old IBMs have recently been discounted. For very simple applications needing no graphics or high performance, a PC or XT or a clone will suffice. For graphics, large databases, word processing with choices of character size and style, and extensive spreadsheets, a PC or XT could be slow, but a clone or an AT would be quite adequate.
Remember that many computers are sold like automobiles, with a number of options. Hence you must be sure a price comparison covers equally equipped units. Let's take a look at some approximate prices in September 1987 newspaper and computer magazine advertisements for a microcomputer, one 360 Kb disk drive (or 1.2 Mb in the case of the AT), 20 Mb hard disk, monochrome graphics/text display, monochrome monitor, 640 Kb random access memory, and one serial and one parallel port:
IBM XT, $1,800-plus; brandname XT clone, $1,200; nonbrand-name XT clone, $900.
IBM AT, $3,200; brand-name AT clone, $2,300; non-brandname AT clone, $1 ,800.
Color costs $400 to $800 more. Most non-IBM accessories, such as color boards (EGA or enhanced graphics adaptor), additional memory, and drives, are either technically superior to or far less expensive than comparable IBM products. Very few users now buy a 100 per cent IBM system; for some applications, it is impossible to do so.
Last December, I purchased an AT clone assembled by a technician in my community. It had a 1.2 Mb drive and a 360 Kb drive (to transfer data to other computers), a 30 Mb hard disk, 640 Kb RAM, one serial and one parallel port, an EGA color board, a color monitor, and speed of 6 to 10 MHz-all of which now costs about $2,000. During the first week of use, I discovered the graphics would not work. The technician immediately exchanged the clone for one that did work, and I have since had no problems with the system.
* Apple Macintosh II. Now, what about Apple versus IBM? The controversy over which firm makes the better product goes on. My personal observations will be brief-I can't cover all the issues here that are of potential significance to computer users.
Apple apparently designed a technologically better microcomputer-the new Macintosh 11 with the Motorola 68020 microprocessor-than the old IBMs or the PS/2. Moreover, Apple may be able to update the Macintosh 11 quickly to the already available 68030 microprocessor, which appears to be faster than the lntel 80386. (An Intel 80486 microprocessor is under development.)
Since Apple clones are almost nonexistent and there are fewer differences among Apple computers, installation of accessories is simpler than with IBMs and their clones, and the likelihood of incompatibility is very small. Many IBM PC accessories, particularly less expensive brands, come with manuals full of random thoughts and references to nonexistent switches. Of course, you can call the manufacturer and get *instructions that are understandable if you have a Ph.D. in electrical engineering.
On the negative side, while an Apple may be easier to learn at first, it often will run slower than an IBM when both are in the hands of experienced operators, and it has less capacity. For similar performance, Apple hardware also seems to be more expensive. Accessories, such as additional memory, hard disks, and color, are either not available or usually costlier than those for IBM microcomputers.
There are Apple versions of most types of IBM PC software but not as many specialized programs for scientific applications, medicine, law, and certain commercial areas, such as taxes. IBM users have a wider range of- software options at a lower cost and opportunities to acquire excellent public domain programs.
Therefore, for the majority of scientific and business applications today, I recommend an XT clone or an AT clone. If all the work you want to do can be performed with an Apple, and you do not mind the higher price, the Macintosh 11 may be an excellent choice.
If you buy an IBM clone for $1,000 and later decide to switch to an AT or PS/2, most of your data and software will be reusable, probably without modification. On the other hand, if you switch to an Apple computer, you will have to transfer your data, and some portions of the data may have to be adapted to the new software. Your old IBM programs will not be usable on the Apple.
1. Siguel, E.N. Why the IBM PC? MLO 17(10): 97-98, October 1985
2. Siguel, E.N. How to become IBM compatible. MLO 17(11): 97-98, November 1985
3. Siguel, E.N. A shopping list for microcomputer hardware, MLO 18(9): 83-84, September 1986.
The author is an assistant professor in the department of community medicine at Tufts University, Boston.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||for laboratory use|
|Author:||Siguel, Edward N.|
|Publication:||Medical Laboratory Observer|
|Date:||Nov 1, 1987|
|Previous Article:||Starting a pharmacokinetic dosing service.|
|Next Article:||U.S. blood supply market approaching $2 billion.|