Death of the PC

Written December 27, 2013

The discussion about the death of the desktop has been raging for a while now. In fact in 2010 the death of the PC was predicted to have occurred by 2013. It is now the end of 2013 and a new timeline will have to be declared. There still remains plenty of doom and gloom but I doubt the PC will become extinct anytime soon. What we are seeing is an evolution to small form factors.

Single purpose machine used to take up an entire room and even grew to entire basements as they became more complex. They were replaced by cabinet size machines, then table sized then eventually the ‘big iron’ that now sits on our desk as desktop PCs. The laptop was supposed to be the first real threat to the desktop, then the ultra-thins and minis now we have tablets and cellphones leading the charge. Will we see the death of the PC? I believe that we will in some way, but not right now.

In the first instance we need to consider need, then processing power and finally technology. For the last twenty years people have being buying new computers that were in many instances simply not needed. Most people use their computers for internet connectivity, document reading and typing and a significant minority uses their computers for entertainment. With the exception of this minority, a computer from 1997 could still accomplish the job, slowly, yes, but definitely accomplish the job. Examine the system requirements for Windows XP, which is still a fully functional OS, and the answer to whether a computer from 1997 could meet the needs of most individuals become clear. Even Windows 8 only needs a 1GHz CPU to be installed and we have had those since 1999. The needs of the general public therefore can be taken care of with new software by machines that are up to 14 years old.

The processing power of computers has hit a plateau with only very marginal gains. Only ARM seems to be in a MHz race currently and this seems to be a marketing ploy by manufacturing companies (reminiscent of theIntel vs. AMD MHz race) rather than by ARM company. In fact both AMD and Intel have given up on this ‘arms’ race a long time ago. Intel currently commands the performance per MHz/GHz at a ridiculous price while AMD commands the performance per dollar at high MHz/GHz. Notwithstanding the perceived performance loss: I had a Sempron le-1300 in 2008 that was handling; movies via TV tuner, between 8 and 15 tabs in FireFoxwhile editing photos in GIMP. There has been enough processing power around for a while.
The last two real big gains in computing have been by AMD, x64 and multicore technologies. Since 2003 we have had both and Intel and AMD have been focusing on other things. For example, power savings of the Haswell line is remarkable while AMDs GPUs are unrivaled. In fact, it seems that both Intel and AMD are focusing on different long term goals, Intel wants to cement itself as the fastest and most efficient CPUs while AMD seems to be after the enthusiast, gamers, entertainment  and general use market; the problem is that there is significant overlap among those groups and “fanboism” remains rife. So what can be used to explain the drop in the CPU market?

Upgrades? Oh yeah you heard me, upgrades. When software is upgraded and the machine works the problem is solved that’s a large part of the market right there. In some cases the memory gets upgraded and whoop there it is! Machines that are up to 14 years old capable of getting the job done is the reason for poor computer sales. The fact that many people didn’t really need a computer in the first place and a tablet will do, is part of the drop in sales but real work still requires a proper keyboard and a monitor at least 15″, any smaller and ergonomics become a well understood word to the user so tablets alone cannot explain the drop in sales.

With Android proving that free software is a viable alternative, many people are slowly experimenting with Linux and some are liking it. This idea about a learning curve for Linux died with the popularity of MACs and got a further death knell with windows Vista, Windows 7 was the nail in the coffin and Windows 8 the memorial service. The bottom line is that people realize that they must re-learn the software with each upgrade or change of OS, what some are finding out is that across the board Linux upgrades might maintain a greater level of consistency with the previous version than any Windows OS might. But, that discussion is a whole other debate about OS and their changes.

–Wilton Mcghee