With all the recent discussions of future consoles from Sony and Microsoft, we've decided to tackle one of the most significant questions--should future consoles be upgradeable?
Historically, console add-ons that boosted the performance of the primary unit haven't done well. Sega's Mega-CD and 32x additions to the original Genesis both failed; Nintendo's
attempt to boost the performance of the N64 through the use of a RAM expansion pack saw only a limited amount of success. Any attempt to upgrade a system's core performance risks bifurcating the user base and increases the amount of work developers must do to ensure that a game runs smoothly on both original and upgraded systems.
Sega's 32X expansion for the Genesis, shown against an equally popular product
The other reason hardware developers are leery of upgrades is that a number of games rely on very specific hardware characteristics to ensure proper operation. In a PC, swapping a CPU with 256K of L2 for a chip with 512K of L2 is a non-issue assuming proper platform support -- existing software will automatically take advantage of the additional cache. The Xbox 360
, on the other hand, allows programmers to lock specific cache blocks and use them for storing data from particular threads. In that case, expanding the amount of L2 cache risks breaking previous games because it changes the range of available cache addresses.
The upgradeable nature of PCs is something we pay for in terms of how effectively programmers are able to use the graphics and CPU hardware they have access to.
On the Other Hand...
The other side of this argument is that the Xbox 360 has been upgraded more effectively than any previous console; current high-end versions ship with more than 10x the storage of the original, as well as support for HDMI and integrated WiFi. The PS3 doesn't entirely qualify -- the current version has the dubious distinction of offering fewer features than it did at launch day. Clearly, however, both Microsoft and Sony are interested in developing peripherals and projects that extend the lifetime of their flagship products.
It would also forestall the decline in comparative visuals. Here's Battlefield 3
on the PC as compared to the PS3: (PS3 on top)
Battlefield 3 - PS3 version
Battlefield 3 - PC version
The PS3 version of the game looks like it's being lit by Spike Lee, with high-powered flood lights just out of view. Textures are minimal; the concrete wall at the left might have been drawn using MS Paint.
The differences between the Xbox 360 version and the PC version of Skyrim are even more stark. The Xbox 360 is on the left; PC on the right.
Click to Enlarge - Xbox 360 - Left, PC - Right
We can only feel sorry for the artists, who labored desperately to make Skryim look like Morrowind; lest they be transferred to the studio handling the motion capture for Rosie O'Donnell's upcoming workout game.
The difference between PCs and the Xbox / PS3 are becoming significant enough to impact buying patterns, and neither MS or Sony are likely to have a console ready for debut before the middle of 2013 at the earliest. Instead of leaping back on the same old bandwagon, a little forward thinking would open up new opportunities. In 2005, there were no external interfaces capable of driving a faster GPU. Today, External PCI-E and Intel's Thunderbolt are both fast enough to drive a discrete GPU, while USB 3.0 could be used to connect high-speed external storage. Microsoft could conceivably design an Xbox 720 that offered WiDi as an option for wireless displays rather than relying on traditional cables.
Building even a modest upgrade path into next-gen consoles could give Microsoft and Sony an even longer cycle and a better chance to recoup initial investments by keeping the platform fresh. Whether or not the companies will do so is an open question.