IEEE Spectrum November, 2014 - 32

reflectioNS_BY roBert W. lUcKY

The PC PlaTeau
Has Moore's Law become moot for PCs?
Desktop computers are no longer the polestar of
the computing firmament, as people have embraced smartphones
and tablets by the million. These portable devices have the obvious advantage of being with you wherever you go, but they are
also empowered with a multitude of sensors-such as barometers, GPS, and
accelerometers-that make no sense in a stationary PC. * Still, when at home
or the office, I feel more comfortable interfacing with a physical keyboard,
a large display, a powerful processor, and lots of memory. So, bucking the
trend, I recently invested in a new PC. My existing PC was four and a half years
old and was a packaged system from a large electronics retailer. I assembled
the new PC myself from premium components-CPU, motherboard, memory, power supply, case, and so forth. After I got the new PC going, I studied the old and new systems and thought about what the comparison told
me about the future of the PC. * My first thought is one of admiration for
the design engineers who make heroic efforts to keep up with Moore's
Law. Between the construction times of the two systems, Moore's Law
predicted there would be three doublings of performance, for a potential gain of a factor of eight. But as we know, raw machine speeds have not
been increasing at that rate. In this instance, the older processor clocks
at 2.93 gigahertz and the new one at 3.5 GHz. That leaves a lot of ground
to be made up through architectural improvements. The new processor
has four cores versus two cores in the older one. The new processor uses
22-nanometer lithography, compared with 32 nm in the older. Combined with
a considerably larger die size, this means it has about 1.4 billion transistors
versus a mere 393 million in the older processor. In addition to processor
32

Nov 2014

North americaN

SPectrUm.ieee.orG

oPiNioN

improvements, significant improvement
in peripheral and storage access speeds
has been afforded by new standards in
USB 3 and SATA 3.
Depending on which benchmark programs I run, the actual observed factor of improvement in performance
between the old and new systems varies from about two to six. However,
this brings me to my second thought:
Unless I'm running benchmark programs, I don't notice any difference in
performance. Most everything I do is
either so fast anyway that I don't notice
a difference or is constrained by some
other bottleneck, like Internet connectivity. Of course, if I were a heavy user
of games, the difference might be evident, but it would depend mostly on
the graphics card used.
When I examine the two systems visually, I'm struck by another thought. Inside
the case, the older, packaged system
looks cheap. The power supply is minimal, the connecting cables hang loose,
and the motherboard looks flimsy and
is populated with no-name components.
Nonetheless, this system has worked perfectly for almost five years, and I bought
it at a bargain. Perhaps there are certain
virtues in cheap electronics in a throwaway era. Maybe I didn't really need all
those premium components in the new
system after all.
My final thought is one of nostalgia
and regret. I think of my PC as a machine and those other computational
devices as gadgets. Moreover, the PC is
the last vestige of visible, accessible electronics in the home. Those other gadgets don't even have any insides-at least
not any I can get at, as not so much
as the head of a screw mars their contoured casings. But, alas, with diminishing incentives to replace older PCs
and the increasing power and ubiquity
of smartphones and tablets, it's hard to
imagine a bright future for the machine
that, starting with the Altair 8800 in
1975, so changed the world. n
photo-illustration by

Gluekit

http://SPectrUm.ieee.orG

Table of Contents for the Digital Edition of IEEE Spectrum November, 2014