The PC era, Apple, IBM, Microsoft, and the home computer

Field note. The reason a game server can run on a $200/mo machine instead of a $200K mainframe is this exact 1980s collapse of the price-to-power ratio. Hosting as a business only exists because of it.

For most of computing's history, computers were big, expensive, and used by institutions. By the late 1970s, that started to change. By the late 1980s, computers were on desks in offices and homes. By the late 1990s, they were everywhere.

The transition wasn't preordained. It was driven by specific decisions, by specific companies, in a specific sequence. This article tells the story.

Before the PC

Through the 1960s and 1970s, "computers" meant:

• Mainframes (IBM, DEC). Million-dollar machines shared by hundreds of users via terminals.
• Minicomputers (DEC PDP, others). Hundred-thousand-dollar machines used by departments.
• Workstations (Sun, Apollo, later). Tens of thousands of dollars, used by single engineers.

Each was professionally managed. Programs were carefully maintained. End users typed at terminals connected to a central machine.

The idea that an individual would own a computer at home was niche. Hobbyist computers existed (Altair 8800 from 1975, others), but they were kits, required electronics expertise, and did relatively little.

The Apple II and the consumer foothold

In 1977, the Apple II launched. It was complete: keyboard, monitor (sold separately), built-in BASIC, color graphics. You took it out of the box and used it. No soldering required.

Steve Wozniak designed the hardware. Steve Jobs handled marketing. The Apple II sold for $1,300 (about $6,500 in 2026 dollars). Not cheap, but accessible to enthusiasts and small businesses.

The killer app emerged in 1979: VisiCalc, the first spreadsheet. Suddenly, the Apple II was useful for actual business work. Accountants and financial analysts bought Apple IIs specifically for VisiCalc.

Apple II remained in production until 1993, an extraordinarily long run. It established that personal computers had real markets.

The IBM PC (1981)

IBM, the dominant mainframe maker, watched the personal-computer market with concern. In 1980, they assembled a small team in Boca Raton to build an IBM personal computer quickly. To move fast, the team used off-the-shelf parts and licensed software rather than building everything in-house.

The IBM PC launched August 12, 1981:

• Intel 8088 CPU.
• 16 KB of RAM (expandable).
• 160 KB floppy disks.
• BASIC ROM from Microsoft.
• Optional MS-DOS operating system, also from Microsoft.

Price: $1,565 base, $3,000 or more with the full package.

The decisive choice: IBM did not lock down the design. The architecture was published. Other companies could (and did) build IBM-compatible PCs ("PC clones"). Compaq released the first major clone in 1983.

This openness was IBM's biggest gift to the world and its biggest strategic mistake. The PC standard exploded as a market because anyone could build PCs. IBM lost control of the market they created.

MS-DOS and the Microsoft pivot

Microsoft was a small company in 1980 (~30 employees), best known for BASIC interpreters. IBM approached them for the PC's BASIC, then asked about an operating system. Microsoft didn't have one but bought QDOS (Quick and Dirty Operating System) from Seattle Computer Products for $75,000.

Microsoft rebranded it as MS-DOS and licensed it to IBM (and crucially, to clone makers) at low per-copy cost. Every PC sold ran MS-DOS. Microsoft's revenue scaled with PC sales.

By the mid-1980s, MS-DOS was on millions of computers worldwide. Microsoft had become the dominant software company of the PC era through one excellent contract decision.

The Mac, the GUI, and Apple's parallel path

Meanwhile, Apple was also evolving. In 1979, Steve Jobs visited Xerox PARC and saw the Xerox Alto, a workstation with a graphical user interface (GUI), bitmap display, mouse, and overlapping windows.

This was a radical departure from text-based interfaces. Jobs returned to Apple convinced that the future was GUI.

The Apple Lisa (1983) was Apple's first GUI commercial product. It failed: $10,000, slow, limited software.

The Macintosh (1984) was the redo. $2,500, faster, with a clearer GUI, and aggressive marketing (the famous "1984" Super Bowl ad). It was a hit.

Macintosh established that:

• GUIs were viable for personal computers.
• Apple's premium-priced, design-focused approach could compete with IBM-compatible volume.
• Polish and integration matter.

Apple's market share peaked at about 10 percent of PC sales in the mid-1980s, then declined as Windows-based PCs caught up. But the Mac established a parallel path that has been important ever since.

Windows takes the GUI mainstream

Microsoft watched the Macintosh and recognized GUIs were the future. They worked on a graphical layer for MS-DOS:

• Windows 1.0 (1985): rudimentary, didn't sell.
• Windows 2.0 (1987): better, still marginal.
• Windows 3.0 (1990): finally good. Sold millions.
• Windows 3.1 (1992): the version everyone remembers from early 1990s offices.
• Windows 95 (1995): a major release. Brought Mac-like usability with PC compatibility. The cultural moment of the PC era.

Windows 95's launch (August 24, 1995) was a global event: midnight openings, TV ads featuring "Start Me Up" by The Rolling Stones, lines at computer stores. Operating system launches had never been (and largely haven't been) like this since.

By the late 1990s, Windows was on ~90 percent of all PCs. Combined with MS-DOS's earlier dominance, Microsoft had been the operating system of the PC era for almost 20 years.

The Office suite

Microsoft Office (Word, Excel, PowerPoint) became the application standard:

• Excel (originally for Mac in 1985, Windows version 1987) eventually displaced Lotus 1-2-3.
• Word (1989 on Windows) displaced WordPerfect.
• PowerPoint (acquired 1987) became universal.

By the mid-1990s, Office was the most-used software in offices worldwide. Microsoft's revenue from Office was eventually as large as from Windows.

Office plus Windows plus the Server products gave Microsoft enormous corporate-IT presence that lasted into the 2010s.

The home computer revolution

The early 1980s also saw a wave of cheaper home computers:

• Commodore 64 (1982): $595, sold ~17 million units. The best-selling single computer model ever for a long time.
• Atari 800.
• TI-99/4A.
• Sinclair ZX Spectrum (UK).
• BBC Micro (UK, in schools).
• Amiga 1000 (1985): multimedia-focused.

These competed with IBM-compatible PCs at the consumer end. For a few years in the early 1980s, "Commodore vs Apple vs IBM PC" was a real debate.

By the late 1980s, the IBM-compatible (with Windows) had won. The home-computer brands either pivoted or died. Commodore went bankrupt in 1994. Atari faded. Apple barely survived to become the company we know.

The PC ecosystem

What the PC era produced:

Standardized hardware. Buy a part, plug it in, mostly works.

Software ecosystem. Tens of thousands of applications across categories.

A computer industry workforce. Programmers, system administrators, support engineers, IT departments. Roles that didn't exist in 1975 became normal by 1995.

Consumer relationship with computing. Hundreds of millions of people who had never used computers before learned to.

Economic transformation. Productivity gains, new business categories (desktop publishing, computer-aided design, accounting software).

By 2000, the PC era's biggest products were:

• Microsoft Windows + Office.
• Intel CPUs (most PCs ran on Intel x86).
• Major software brands (Oracle, SAP, Adobe).

The decline and what came after

The PC era's peak was approximately 1995-2005. Then:

The internet became dominant. Computing's center of gravity shifted from "individual desktops with installed software" to "web-based services."

Mobile devices. Starting 2007, smartphones became the primary computing devices for most users. PCs receded into being work-tools.

Cloud. Application logic moved to servers; PCs became thinner clients.

Apple's resurgence. Macs grew share. The iPod, iPhone, iPad shifted Apple's center of gravity from PCs to mobile-and-services.

PCs still exist. Hundreds of millions of new PCs ship every year. But the era when "buying a PC" was the focal moment of consumer technology ended around 2010-2012.

What the PC era proved

A few lessons:

Open standards win mass markets. The PC's open architecture enabled the clone industry and accelerated adoption.

Software lock-in is real. Microsoft's Windows-and-Office combination kept users on Windows for decades through application compatibility.

Distribution matters. Bundling Windows with PCs and Office with Windows gave Microsoft enormous advantages.

Design matters too. Apple's premium-design approach maintained a viable parallel market even when their share was small.

Markets transform over decades. The PC era took 15-20 years to develop and another 15 years to fade. Tech transitions are slower than they feel.

The PC's enduring presence

In 2026, PCs are still important:

• ~250 million new PCs sold per year.
• Most office work happens on PCs (or Macs, treated as PC-equivalent).
• Software development is mostly done on PCs or Macs.
• Gaming is partly PC.

But PCs are no longer the focal point of consumer technology. They're the productivity workhorses; mobile is the social/consumption platform.

Conclusion

The PC era took computing from rooms to desks to homes. It was driven by a few specific products (Apple II, IBM PC, Macintosh, Windows) and a few specific companies (Apple, IBM, Microsoft, Intel).

By 2000, computing was a consumer product, not just a professional tool. Hundreds of millions of people had personal computing relationships. The internet, when it became dominant, found PCs ready to receive it.

For the modern internet era (2000+), the PC era was the foundation. Without millions of PCs already in homes, the Web's adoption would have been much slower. The two transformations built on each other.

Coming up

Next: Linus Torvalds and Linux. A 21-year-old Finnish student decided to write his own UNIX kernel as a hobby. It became the most-deployed operating system in history.