Retro Computing

Staff writers | May 17, 2007 1:38 PM
PC Authority takes a trip down computer memory lane and discovers just how PC technology has evolved... and how Microsoft and IBM nearly destroyed each other.
Way back in November 1994, we were worried about how our PCs would cope with the new-fangled plug-and-play features in the forthcoming Windows 95, marvelling at how thousands of new Internet users signed up every month, and praising those PCs that included a CD drive.

Jump forward to the present, and we’re wondering how our PCs might cope with Vista’s 3D user interface and lavishing praise on laptop makers for bundling HD DVD drives. The dizzying speed of progress in the PC industry is astonishing, as you’ll see over the following pages, where we benchmark a classic 486 system against a cutting-edge powerhouse, and spend a fortnight working with Windows 95 in a bid to explode the myth that nothing much has changed in PC software since the mid-1990s.

However, our feature isn’t only a nostalgia-filled view of life from the days when 56K modems were being touted on shows like Beyond2000: we also canvas the opinions of some of the industry’s brightest minds to find out what we can expect to see over the next couple of decades. If we’re wondering how we ever survived without our quantum computing robots in ten year’s time, remember that you read it here first. We examine how PCs have evolved since PC Authority began, and put a vintage machine through its paces by feeding it the application tasks of today.

Looking back over the past decade, you can see the rough-and-ready interpretation of processor speeds as indicated by FLOPS. This is only a basic indication, based on maximum theoretical floating-point operations per second. It’s a purely theoretical number based on the notion that the floating-point units can complete a certain number of operations per clock cycle. For early processor designs, this was simply one operation per clock, hence a 333MHz Pentium II could, in theory, manage 333MFLOPS (a MFLOPS is one million FLOPS).

Real performance benchmarks, on the other hand, are based on hundreds of other factors. You’ll notice that the theoretical MFLOPS performance didn’t alter between our average Labs-test PCs in a two month period, because both had 3GHz Pentium 4 processors. The maximum theoretical floating-point performance of any Pentium 4 is simply the maximum number of floating-point operations per clock, multiplied by its clock frequency; the 128-bit SSE registers in a Pentium 4 allow four floating-point operations in a single clock cycle in the ideal case, making for a maximum FLOPS rating of 12,000MFLOPS (or 12GFLOPS). But in fact the Pentium 4 variant in our later tests was a superior part for real-world applications, having more cache memory and Intel’s HyperThreading system, which presented the system with a second virtual processor constructed from idle resources. In practice, of course, real-world performance is about the complete PC, which is what PC Authority’s application-based benchmarks have always been about.

Application benchmarks are the only way to really measure true performance, especially today. Most architectural enhancements in processors – such as branch prediction, out-of-order execution and speculative instruction fetch – rely on the typical non-linear nature of applications. Feeding them synthetic tasks that simply repeat the same tight loop over and over again doesn’t simulate that.

Second, of course, the processor is only one of half a dozen or so key components that affect performance. The speed of ancillary components is more important in the real world for many apps – hard disk speed chief among them – and the bandwidth of internal buses and interconnects is a major bottleneck when it comes to shuffling data from processor to memory and graphics card. With that in mind, how does a machine from several years ago fare against one of today?

In context
Back in 1994, the PCI bus was the height of new technology. Running at 33MHz with a 32-bit bus width to the ageing ISA bus’ lowly 8MHz at 16 bits, it could push around 125MB/s between a graphics card and main memory. The (now obsolete) AGP port wouldn’t be invented for another three years, and the idea of a serial interface for anything that needed seriously fast transfer rates was pure fantasy; parallel buses were clearly superior, since you could push a couple of bytes per clock down the pipe as opposed to just one measly bit. 1994 hard disks were connected with an early version of the IDE (integrated drive electronics) interface. With a bandwidth of 33MB/s, it was more than sufficient, since only the fastest hard disks could approach10MB/sec transfer rates. Incidentally, hard disks were all formatted with the FAT16 filing system, meaning you couldn’t have a single partition larger than 2GB. This wasn’t a terrible burden, as the largest hard disks were only around 540MB anyway.

These days, the serial-transfer paradigm has sensibly taken over from the old parallel way of doing things, and bandwidth is frighteningly high. Issues of cross-talk and clock-skew meant that parallel interfaces had hit the buffers and extremely high-speed serial interfaces are now the way forward. Hence, we have SATA and PCI Express, both based on the same fundamental high-speed serial transfer technology. Where the PCI bus of 1994 could transfer 33MB/s, the PCI Express interface of today can push 8GB of data from north bridge to graphics card every second.

And where the PATA-based IDE hard disk interface of 1994 topped out at 33MB/s, SATA can now shove 300MB/s down the pipe. This is just as well, since a single high-end hard disk such as Seagate’s 15K Cheetah can now achieve almost 100MB/s, and formerly stupendously expensive RAID multidisk capabilities are routinely built into motherboards costing $80. Just hook up a few consumer-level drives costing $150 or so and you’ve got 100MB/s, no problem.Testing the old guard
It’s saying something in itself that it’s simply impossible to test a 486 system, circa 1994, in a meaningful way against a modern PC. With an average complement of 8MB of RAM against the 2GB of today, you couldn’t even begin to load up Windows XP. For a bit of fun, try right-clicking on a clear area of your Taskbar and selecting Task Manager. Hit the Processes tab and then click on the Mem Usage column to order your running programs by the amount of memory they’re consuming. When we did this, we found that a single instance of Firefox was using 56MB – seven times the total memory complement of a 486 PC. Windows Explorer was using 47MB and the Word software we used to write these very words had 27MB all to itself. Enough said.

But by December 1998, PCs had reached an average of 80MB of RAM – in our Labs group test, the entrants were a mix of 64MB and 128MB systems. Average hard disk capacity was around the 6GB mark (the FAT32 filing system allowing larger partitions) and the favoured processor was a 333MHz Pentium II.

This is the point at which it becomes possible to at least attempt to load up Windows XP and run our present-day benchmarks on an old system. So that’s exactly what we did.



The results
Remember that in our current benchmark suite, the results are normalised to our 3GHz Pentium D reference system. This in itself is getting long in the tooth, and new PCs now routinely score over 2.00, meaning they’re twice as fast as the reference. The front-end application to our benchmarks suite, understandably, doesn’t normally expect to see scores of less than 0.1 and got a little grumpy, so we resorted to trickery. We managed to get both our 3ds Max and Photoshop tests going, although the rest of our tests proved too much for the 128MB of RAM on offer. These two are actually good tests to run, though, since 3ds Max is almost completely CPU-bound, whereas Photoshop is a memory hog, thrashing the hard disk. It turns out that a PC from today’s testing is on average about 50 times faster than one from the middle of our magazine’s history; a frame of our 3ds Max test completes in an achingly slow 28 minutes, against the 50 seconds or so of a modern PC. And Photoshop sits and grinds for over four hours before
completing a single run of our tests, compared to the five-and-a-half minutes or so of a new machine. That, ladies and gentlemen, is progress.Has software’s progress really been just pretty add-ons and superfluous upgrades?

Many people think PCs haven’t evolved much since Windows 95 sprang onto the scene with its groundbreaking Start button, My Computer and My Documents facilities. Are the critics right, or has software’s gradual upgrade path made us blind to the changes over time? To find out, I reluctantly traded in my home and work PCs for two old Pentium/60s, both replete with 16 whole megabytes of RAM and 850MB hard disks, and attempted to install Windows 95 on both of them. Would it be possible to go about my computing life and work as normal, or would I be screaming for my regular PCs within hours? After starting the Windows setup process with a DOS boot disk, I then had to set up the hard drive, and the only file system available to me was FAT. NTFS had yet to become mainstream in 1995, and even good-old FAT32 was out of the question until Windows 95b rolled along. This meant I had to use an ageing 16-bit file system, despite having a 32-bit OS. Not only that, but it also limited the size of a hard disk partition to 2GB,although this wasn’t a problem with my dinky 850MB disks.

Windows 95 couldn't even set the wallpaper right.
Windows 95 couldn't even set the wallpaper right.


That’s not to say the Windows 95 installation went entirely smoothly.
For example, it would often say “Please insert the Windows 95 CD-ROM,” even though the CD was already in the drive. This foible worsened when I tried to install the graphics drivers. Windows prompted me for the driver’s location, so I browsed to its folder on the hard drive. Is that where it looked? Of course not. The floppy drive clicked and I was told that Windows couldn’t find the driver. I then had to browse to the driver’s location again. Windows found it this time, and then asked me to insert the Windows 95 CD-ROM. This looked promising; it had the letter of the CD-ROM and the directory of the file in its Location box, but when I inserted the CD and clicked OK, the PC again said that it couldn’t find it. I then had to browse through all of the directories on the CD just to find the blasted file. Windows 95 was far from a finished product when it hit the shelves, and I’m not only talking about the installation and setup process. Microsoft couldn’t even get trivial embellishments such as the wallpaper right: Windows 95 would only display a wallpaper image at its actual size in the middle of the screen, or tiled. A tiled background of tessellated patterns looks alright; a tiled photo of the Windows’ clouds looks stupid. Even more infuriating was the way Windows 95 displayed files and folders. For example, if you open My Documents or Control Panel and adjust the size of the window, you’d expect the icons to move around to fit inside your newly sized window. Not so with Windows 95; they stay exactly where they are and create big scrollbars in the process. Lunacy. Reorganising the Office
The next step was to install Office 95 – no mean feat when you consider the beast arrived on 24 floppy disks. I’d bought the disks second-hand and, as the disks were writeable, it knew the software had been installed more than three times. This meant, in addition to the disk-swapping marathon, I had to click OK on every single file in the installation process, just to confirm I wasn’t a filthy software thief. The whole process took a painful hour to complete. That said, when Office was installed, my working life was very similar to how it is with Office XP. Of course, much of this is down to how I use Word and Excel. Many people I know couldn’t work without the annotation features found in Office XP onwards, but like most people I just use Word for basic documents, for which Word 95 is perfectly sufficient (although it’s worth noting that the eccentric grammar-checking feature has been notably improved). My next challenge was to connect to the company network, and I was surprised to find that TCP/IP wasn’t installed by default. Indeed, the only networking protocols available as standard were IPX and NetBEUI. This wasn’t a big problem, though; I just had to install the TCP/IP protocol, and then type in my IP address and subnet mask. This project set off the alarm bells in our IT department, though, as the concept of a firewall was largely unheard of back in 1995, and any decent security software now only works with Windows 2000 and XP. Undeterred, I carried on, and the PC logged onto the firm’s network with no problems. When I got home, I then had to set up my 14.4K modem, which was amazingly pain-free. I did, however, thank my lucky stars that I was dealing with Windows 95 rather than Windows 3.1, as setting up dial-up networking on the latter was akin to fitting a horse in a CD rack. Nevertheless, accessing the Internet back then wasn’t the cakewalk it is today. Let’s start with email: bearing in mind that Windows 95 predates Outlook Express and even Microsoft’s Internet Mail app, you were stuck with the primitive and unfriendly Windows Messaging Service (unless you opted for third-party software). It starts by asking for the address of your Post Office – whatever that was – and then proceeded to ask for a Mailbox name that could only have eight characters. Confused, I then discovered that, like TCP/IP, Internet Mail support wasn’t installed by default, so I had to go into Add/Remove Programs, click the Windows Setup tab and install it separately. By comparison, both mail and TCP/IP support are both a given now. This was merely the tip of the iceberg, because next came web browsing courtesy of Netscape 2. Of course, this is one area of testing that could never be completely accurate. I didn’t have a time machine to take me back to 1995, when all websites had the same lamentable HTML standards and usability as a goth kid’s MySpace page. Surprisingly, though, the experience proved to be very revealing, even with today’s sites. Netscape 2 knows nothing of Flash or Shockwave, and it displays today’s sites much as they would have looked 12 years ago. The browser accordingly put grey borders around frames, downloaded each image a bit at a time (slowly), and very often couldn’t find the images and brought up the highly nostalgic coloured shapes in their place. I was instantly reminded of just how awful the Internet was in 1995.

376MB is all you need for basic day-to-day work, but you can forget having any fun with an 850MB hard drive.
376MB is all you need for basic day-to-day work, but you can forget having any fun with an 850MB hard drive.


Where’s my features?
Nowadays, people think it’s strange that Windows Media Player will only play DVD movies if you have a third-party playback codec installed, but that’s nothing compared to the barren desert of software supplied with Windows 95. I sorely missed my MSN Messenger (although there were IM alternatives at the time). With Windows 95, you just get a joke of a music player, which can play only WAV, MIDI and CD audio files, along with MPG and AVI movie files. Anything more than that was asking too much. But we didn’t need to play back much else back then, as MP3 music had yet to become the massive craze it is now. I wouldn’t have wanted to download loads of MP3s anyway, as that would have swamped my 850MB hard drive in no time and CD burning required third-party software. Strangely, though, the hard drive never swelled beyond half-full in my tests. It may well have done if I’d filled it with games and other apps, but I found that the 376MB I used provided ample space for day-to-day life. That’s not to say we don’t need
large hard drives now; it just means our needs have changed. But, although I could survive happily with my minimal hard disk, the meagre RAM really affected my work. I’m now quite used to having a browser with several tabs open simultaneously, as well as Word, Outlook and Photoshop open at the same time, but this proved to be too much for my ancient machines with their pitiful 16MB of memory. I could just about have Word and Netscape open at the same time, but I was harshly punished whenever I double-clicked on an application icon by mistake and had to wait for the program to load, just so I could close it. There were also numerous occasions when apps overlapped one another on the screen, as Windows hadn’t refreshed itself properly. Similarly, I also noticed how reliant we’ve become on USB. The only form of removable storage in my PC was the unreliable floppy drive. It’s a pain when you’re used to plugging in a USB drive and getting masses of storage.

Verdict
Before I went back to these old machines, I was one of those people who thought software hadn’t really changed much. But, in the same way your school memories are rose tinted, it turns out that software is a whole lot better than it was when Windows 95 came out. We might not have had the leap in UI design there was between Windows 3.1 and 95, but the software we use now is much more polished, and there’s a hell of a lot our PCs do now that we take for granted. Be thankful for USB flash drives, ubiquitous instant messaging, speedy web browsers and easy Internet setup. They might not seem like a big deal, but they sure are when you have to make do without them.Windows has dominated the pc platform for more than a decade, but how did it achieve its enormous success?

There’s an urban legend that Microsoft Windows came to rule the world by beating Apple’s Mac OS. In fact, Microsoft had seen off Mac OS with MS-DOS long before Windows ever arrived. Nevertheless, Windows did face some serious challengers along the way to world domination. And although there wasn’t a single reason for its eventual triumph, Windows benefited from a tremendous amount of luck. The first example of Microsoft’s good fortune was the company winning the contract to produce the disk operating system for the IBM PC, launched in 1981. That was amazingly lucky, because IBM had intended to use Digital Research’s CP/M. When that deal fell through, it landed in Microsoft’s lap. Even more fortuitously, Microsoft knew someone – Tim Paterson of Seattle Computer Products – who’d already written a basic CP/M clone for the PC’s chip, which it could buy. The operating system was renamed PC-DOS (later to be rebranded MS-DOS) and Microsoft, a tiny 40-person operation with only $7.5 million in revenue, was
on its way.

IBM's OS/2 was Windows' most serious threat.
IBM's OS/2 was Windows' most serious threat.


Going for GUI
Clearly, MS-DOS wasn’t going to last, so the next step was to find a replacement. At the time, everyone and his dog was working on graphical user interfaces (GUIs), inspired by Xerox’s work in the 1970s, and Microsoft was no exception. It announced Windows in 1983 and shipped the first version in 1985. However, Microsoft faced stiff competition for Windows from Digital Research’s CP/M, which dominated the business market. Fortunately for Microsoft, CP/M-86 cost about four times as much: Digital Research believed IBM needed CP/M and that customers would pay a premium for it. It was wrong.

Digital Research wasn’t done, however. The company produced a better graphical front end than Windows, in the form of GEM (Graphical Environment Manager). It also had some reasonable applications, including GEM Write, GEM Paint and Ventura Publisher. And, unlike Windows, it ran on both business-standard Intel x86 chips and the racier Motorola 68000. Game over? No, Microsoft got lucky once more. Redmond’s second biggest software house (after Nintendo) had a big buddy called Apple Computer, which was desperate for applications for its Macintosh computer. Apple didn’t like Digital Research because it had sold GEM to Jack Tramiel for his Atari ST, which InfoWorld magazine called the “Jackintosh”. The ST was slightly faster than the Mac, had a better screen and cost much less. No fly on the wall has reported what really went on, but Apple gave Microsoft rights to the Mac interface and sued Digital Research over the “look and feel” of GEM. Microsoft improved Windows, while Digital Research capitulated and never recovered. (Later, Apple sued Microsoft and lost.)


Battle with Big Blue
One reason why Microsoft was keen on the Mac was because of its problems with IBM. The PC had been developed in a hurry by a “wild duck” group, contrary to IBM’s usual practice of owning the whole computer industry. When the PC succeeded beyond anyone’s wildest dreams, IBM pulled the PC division back into the fold and decided to change the industry to an IBM-owned standard. This resulted in the 1987 launch of the PS/2 (Personal System/2) with its new MCA (Micro Channel Architecture) bus and OS/2 operating system. IBM was going to stop making IBM PCs! IBM, a $54 billion giant, had a whole business strategy called Systems Application Architecture to tie its mainframes, minis and PCs together. This required OS/2 Extended Edition, which was exclusive to IBM. Microsoft, a mere $0.35 billion tiddler, looked doomed. Or was it?

In 1983, Microsoft had signed a joint development deal for OS/2, and it could sell the software to the non-corporate users who weren’t “IBM shops”. Microsoft was hard-core about OS/2. In his foreword to the OS/2 Programmer’s Guide (1988), Bill Gates wrote: “I believe OS/2 is destined to be the most important operating system, and possibly program, of all time.” Unfortunately, its sales were, in Bill’s own words, “dismal”. The 16-bit version, written for Intel’s doomed 286 chip, had no GUI at launch, almost no applications, consumed too many expensive resources and couldn’t multitask DOS programs. Nobody wanted it. And while IBM could afford to keep more than 1,000 programmers hacking away at OS/2, Microsoft could not. Again, the situation looked impossible, but, again, Microsoft had two incredible strokes of fortune. The first was that Digital Equipment Corp (DEC), a minicomputer giant, fell out with its star programmer Dave Cutler, who’d developed VAX VMS. Microsoft told him, in effect: we’ll back you, bring everyone you want. Cutler moved his team to Microsoft in October 1988 and started the development of a new 32-bit cross-platform multiprocessor operating system that would run applications written for OS/2.

Apple rejected Microsoft’s offer to make Mac OS an industry standard.
Apple rejected Microsoft’s offer to make Mac OS an industry standard.


The second incredible piece of luck came when a Microsoft programmer ran into an old friend at a party. The friend, a physics professor who was working on Microsoft’s CodeView debugger, explained how it would be possible to get Windows running in the 386 chip’s protected memory mode. (IBM had limited DOS to 640K, remember?) This magic fix made the successful launch of Windows 3 in 1990 possible. Users no longer had to abandon DOS and change to a new operating system such as OS/2: they could simply add Windows – it was just like installing an application – and use it as and when they liked. There was no risk: DOS was still there and, with Windows, they could run multiple DOS applications.Power shift
Suddenly and quite unexpectedly, Microsoft no longer had to do what IBM said. It started pitching Windows as a stepping stone to OS/2, but, naturally, its growing popularity put the skids under the rival OS. IBM had thrice refused to co-develop and support Windows, snubbed the Windows 3 launch, and now took to demonstrating ways to crash it. Things got nasty. But IBM also had a choice. It could either keep rewriting OS/2 itself or do a deal with Microsoft to adopt the portable OS/2 NT. After some long and bitter arguments, the two companies finalised their divorce in 1992. IBM got all the code for 16-bit OS/2 1.3 and the 32-bit version 2.0, a cheap deal for Windows and $25 million in cash for IBM software patents. Microsoft got a nice royalty on all sales of OS/2 and the code for OS/2 3.0 NT. Microsoft then had to scramble to convert NT from supporting OS/2 applications to supporting Windows applications. IBM moved to total war, trying to kill the upstart traitor; one senior IBM executive told me: “We’re going to burn Bill’s butt”. It spent well over $1 billion on OS/2 programming and hundreds of millions of dollars on marketing. This effort continued until an outsider, Lou Gerstner, arrived as IBM’s new chief executive. In his book about how he saved IBM, Gerstner says he found himself embroiled in a religious war that “we were going to fight to the bitter end. What my colleagues seemed unwilling or unable to accept was that the war was already over and was a resounding defeat.” And what of 1994's OS/2 Warp? According to Gerstner, it was “the last gasp.”

Bill Gates lead Microsoft to victory against IBM; IBM’s Lou Gerstner conceded defeat to Microsoft’s Bill Gates.
Bill Gates lead Microsoft to victory against IBM; IBM’s Lou Gerstner conceded defeat to Microsoft’s Bill Gates.


Microsoft finally won that battle with the launch of Windows 95. That established Windows as an operating system, rather than an optional utility running on top of DOS. After that, Microsoft only had to transition the Windows market from old DOS/Windows to a proper New Technology operating system, NT.

If you’ve just moved to Vista, you’re using an operating system that started in the 1980s as OS/2 NT and went through NT3.5, NT4 and Windows 2000 before it finally took over with Windows XP. The moral of the story? The slow process of incremental evolution is boring, but desirable. Revolutions are exciting, but they leave a lot of dead bodies on the floor.

Unix and Apple
So what happened to Unix? In theory, Unix should have knocked Windows into a cocked hat, because it ran on everything from PCs to supercomputers and wasn’t tied to one proprietary manufacturer. In reality, the dozens of different versions had extensions that made them incompatible – plus, it was expensive, user-hostile and a resource hog. What the industry needed was a single unifying standard, so AT&T and Sun got together behind everyone’s backs and created one – SVR4 (System V Release 4) – to their own advantage. Others rebelled, setting up the Open Software Foundation to develop a rival, OSF/1. (“Oppose Sun Forever,” quipped Sun’s Scott McNealy.) Chaos returned and the industry crumbled. The mutual back-stabbing subsided in 1993 when the two sides joined to form COSE (Common Open Software Environment), but that was too little and far too late. And what of Apple? Did the company really have a chance? Maybe. In 1985, Bill Gates and Jeff Raikes wrote to Apple suggesting it license Mac OS to make it an industry standard, and offered to help. Apple declined because it couldn’t see the sense in selling an operating system for $50 or $100 when it could sell Macs for up to $5,000 each. What Apple also failed to realise, until it was too late, was that buyers would stick with DOS, the industry standard established by IBM, rather than switch to Mac OS. DOS gave them a vast range of applications and the choice of buying from thousands of PC suppliers, rather than locking them to one. Many businesses had been burned when proprietary suppliers of mainframes and minis went bust. That’s why people used to say that no-one got fired for buying IBM. Well, in the PC world, no-one got fired for buying IBM-compatibles. Microsoft’s real triumph was that it was able to nick IBM’s monopoly, converting “IBM-compatible” into “Windows-compatible”. Was that luck or good judgement?

Buyers shunned the proprietary Mac OS.
Buyers shunned the proprietary Mac OS.
This article appeared in the May, 2007 issue of PC Authority.