The IBM Personal Computer was announced on August 12, 1981. Having sold its PC business to Lenovo in 2005, IBM itself did not mark the occasion, and as far as I know, neither did Lenovo or the other companies still selling IBM PC’s. But, the 30th anniversary of this important event in the history of computing was duly noted by IBM Fellow Mark Dean, - a member of the small team that designed the original machine, - in a nostalgic and thoughtful blog. At the beginning, Mark writes:
“It’s amazing to me to think that August 12 marks the 30th anniversary of the IBM Personal Computer. The announcement helped launch a phenomenon that changed the way we work, play and communicate. Little did we expect to create an industry that ultimately peaked at more than 300 million unit sales per year. I’m proud that I was one of a dozen IBM engineers who designed the first machine and was fortunate to have led subsequent IBM PC designs through the 1980s. It may be odd for me to say this, but I’m also proud IBM decided to leave the personal computer business in 2005, selling our PC division to Lenovo. While many in the tech industry questioned IBM’s decision to exit the business at the time, it’s now clear that our company was in the vanguard of the post-PC era.”
“I, personally, have moved beyond the PC as well. My primary computer now is a tablet. When I helped design the PC, I didn’t think I’d live long enough to witness its decline. But, while PCs will continue to be much-used devices, they’re no longer at the leading edge of computing. They’re going the way of the vacuum tube, typewriter, vinyl records, CRT and incandescent light bulbs.”
A few days later, I saw a couple of articles with headlines that read: “The PC is dead, claims IBM chief tech officer on 30th anniversary of home computer launch,” and “IBM Inventor: PC is dead.”
At the time I had not read Mark’s blog, as I was busy getting ready to give a keynote presentation the following week at LinuxCon in Vancouver. I made a mental note to check what Mark had actually written after I returned from Vancouver.
I gave my presentation on August 18 in the morning, right around the time that HP announced that it was considering spinning off its PC business. Later that day, during an interview about Linux and related subjects, a reporter asked me in passing what I thought about the earlier stories that the PC was dead, which were given further prominence given HP’s announcement that morning.
I said that it all depends what you mean by dead. I did not think that personal computers were any more dead than mainframes or supercomputers, which have been around since the early days of the IT industry. While the role of the mainframe has evolved over time, it continues to do quite well. Although no longer at the center of computing, PCs could similarly evolve and do well for years to come.
Having now had more time to think about this question, it might be instructive to look at how mainframes fared in their own post-mainframe evolution as a possible lesson for the future of personal computers.
In the 50 - 55 years since there has been an IT industry, we have had two major eras in computing: the centralized and the client-server eras, with mainframes and PCs as the leading platforms in each respectively. We are now in the early stages of our third, which I think of as the era of cloud computing. No one product can claim to be the leading platform of the cloud era. Rather, that role now belongs to the Internet and everything around it.
From its beginning through the1980s, just about all computing was centralized. The machines were fairly expensive, - mainframes for business applications and supercomputers for scientific ones, - and were generally located behind the glass walls of data centers and shared by many users. Minicomputers were smaller, less expensive and often used in individual departments instead of managed as a central, shared resource. Users interacted directly with the applications on these machines via terminals connected over proprietary networks.
Mainframes were built using fairly sophisticated and expensive technologies. They were designed to be shared by multiple applications and many users simultaneously. Airline reservation, inventory management and banking systems were among their best known applications. A lot of attention had to go into reliability, since a failure would impact all the users and critical applications sharing the computer. The hardware architecture and software were carefully designed so the machines could provide very good response times to high volumes of transactions while operating very efficiently, that is, at high utilizations.
IBM announced its System 360 family of mainframes in April of 1964. Over the next 20 - 25 years, IBM mainframes became the premier platform for commercial computing. A fairly large ecosystem developed around them, with many companies providing a variety of hardware and software add-on products, applications and services.
By the late 1980s, advances in microprocessor technologies were enabling a new generation of vendors to compete with mainframes, supercomputers and minicomputers with significantly less expensive client-server solutions based primarily on PC and workstation platforms. Minicomputers essentially disappeared within a few years, along with a number of the once powerful companies that built them, like DEC and Wang.
Supercomputers went through a major transition. The machines of the 1970s and 1980s, based on specialized vector architectures and expensive, highly sophisticated technologies, were no longer competitive. A new generation of supercomputers emerged based on powerful microprocessors and parallel architectures. Supercomputers have done quite well since, with ever more powerful machines and innovative applications brought to market just about every year.
Mainframes lost the leading position they had previously enjoyed. The new client-server products became the fast growing platforms where most of the innovations now took place. Many predicted the impending death of mainframes. In March of 1991, for example, VC and journalist Stewart Alsop famously said: “I predict that the last mainframe will be unplugged on March 15, 1996.”
IBM went through a very painful near-death experience. Competitors attacked its large mainframe business with less expensive UNIX and PC servers. The company responded by developing a new generation of mainframes based on microprocessors and parallel architectures. The new machines were compatible with the large existing base of mainframe software, and continued to support the high transaction volumes and industrial strength required by mission critical workloads.
Mainframes survived. The architectural descendants of the original System 360 are actually doing quite well. They are no longer the only game in town, but now share the market with a variety of servers from IBM and other vendors. Their development continues. In July of 2010, I attended the announcement of their latest generation. Ten years after his death-of-the-mainframe prediction, Stewart Alsop metaphorically ate his words.
What lessons can we draw from the way mainframes evolved into the post-mainframe world? Mainframes survived because they embraced the changing technologies and innovations that almost killed them. They have kept up with advances in technology, performance and quality over the years. At the same time, mainframes have incorporated major market innovations into the platform, including all the key Internet standards, Linux, Java, Web Services and so on.
Then there is focus. Mainframes have continued to lead in their particular market segment, otherwise customers would have long abandoned them for alternative platforms. Mainframe customers like them because they are a highly cost effective platform for large, complex business applications that require the highest degrees of performance, reliability and security. As long as mainframes can continue to provide good value for such mission critical workloads, customers see little reason to go through the large efforts and costs involved in switching to a different platform.
There might be some good lessons here for personal computers as they transition into the post-PC era, - when they are no longer at the center of computing as they were from the mid 1980s until a few years ago. Personal computers will need to keep incorporating the latest technologies and innovations into the platform, even as many of these capabilities will now be first developed elsewhere. They will have to provide good value for customers, which likely means that prices will continue to drop and business models will be under pressure.
As has been the case with mainframes, focus will be essential. It will be hard for personal computers to still be all things to all people, as they were in their prime. A number of applications once supported on PC platforms have been migrating to different kinds of devices, including smartphones, tablets, e-book readers, music and video players and many new kinds of specialized products designed to support specific apps and industries.
I continue to use my personal computer, for a variety of office and personal productivity applications. But I now also use a tablet and smartphone. Each has a role to play. In my home I have a growing number of specialized, Internet-enabled products, including Tivos, TVs, DVD players, phones and music devices. For me, as for so many others, the center of computing is no longer any one device, but the Internet-enabled clouds out there that they are all connected to and from which we get a large variety of information, apps and services.
So, where is computing heading in the post-PC era?. Let me conclude with the eloquent and succinct answer to this question offered by Mark Dean in his August 10 blog:
“PCs are being replaced at the center of computing not by another type of device - though there’s plenty of excitement about smart phones and tablets - but by new ideas about the role that computing can play in progress. These days, it’s becoming clear that innovation flourishes best not on devices but in the social spaces between them, where people and ideas meet and interact. It is there that computing can have the most powerful impact on economy, society and people’s lives.”