2007-03-14

PC Gaming Dying at 25, Part I: Why?

In this multi-part blog, I will explore the state of PC Gaming after a quarter century, including ...
  • Part I: Why and how is PC gaming dying? (this entry)
  • Looking back at the various genres ... both dead and alive (and sometimes both)
  • The future of smaller commercial and community PC gaming development
Part I: Why and how is PC gaming dying?

The reasons for the slow, but stead death of PC gaming -- at least on a major commercial developer scale -- has everything to do with the set-top console. Specifically, in descending order of relevance from most to least significant ...
  • Hardware entry and development costs v. volume and redesign costs
  • Software entry and development costs v. volume and support costs
  • Consumer preferences of "load and forget" and "reduced limitations"
Hardware entry and development costs v. volume and redesign costs

The first "wake up call" to the changing industry should have been IBM's virtual standstill on the Power4-based PowerPC -- especially the common PowerPC 970 sold under the marketing name G5 by Apple. IBM is the world's largest, independent semiconductor foundry -- i.e., a company that helps layout and actually fabricate large volumes of integrated circuits (IC) including maintaining its own, "core" and "peripheral" set of logic chips -- essentially forced Apple to Intel. Before Apple negotiated its new, preferred status with Intel -- replacing Intel's previous favorite, Dell -- IBM did not deliver on its promises of a low-power PowerPC 970, and failed to ramp up clock and performance on the series.

This had little to do with the actual design, capability and scalability of the PowerPC 970, and everything to do with profit. The PowerPC 970 was a brilliant, powerful, scalable and even very low-power (while still being very fast) PowerPC-compatible, 64-bit Power 4 core. But, as always, economies of scale is everything, especially for a large foundry.

For any given PowerPC 970, IBM did design, layout, tape-out and fabrication of a release of the IC for a 12 to maybe 15 or 18 month (in the best case) run. And during that run, they'd sell maybe 1M+ units to Apple for G5 units (G4 products are Motorola) and, to a far lesser extent, various other IBM microelectronics licensees. In the end, this "hardware entry" was quite steep compared to the "volume pay-off," especially considering the limited lifetime of the design before a revision was required. Hence why IBM eventually reached the real conclusion that a mobile G5 was not really worth their bother.

After all, Motorola had let it's PowerPC (G4 and earlier) line essentially die years earlier, leaving Apple with only IBM. Luckily for Apple, unlike Microsoft, their OS codebase was designed for portability (NT 3.1-4.0 not-withstanding on select Alpha, MIPS and PowerPC, of which the Win32 codebase did not port well because it was mutated by MS-DOS 7+Windows 4 aka Windows 95+, hence dropped as of NT 5.0 aka Windows 2000). It was constantly built and tested and even cross-linked against the most commodity platform on the planet, x86. And to that end, Apple found itself using the ICs from the only one of two companies that ship millions of consumer-grade, desktop/server ICs for general usage with a limited 12-18 month revision cycle, Intel.

So what does this all have to do with PC gaming?

IBM is a foundry. Power, not PowerPC, is their "core" offering (and they have even sold exclusive licenses of PowerPC lines to select companies, like the 400 to AMCC). They have countless design capabilities, plenty of "peripheral" support logic options and many other, unique offerings in their combination of "building block" design and fabrication volume. As such, short of select consortiums (like ARM or MIPS), IBM has a huge market swing when it comes to embedded mindshare. So, why does that matter again?

Imagine you are a design and fabrication firm. Would you like to design (and redesign) a product every 12-18 months, and only see 100,000+ units? Or would you like to design a product once for 6 years, with one die-shrink/ cost-reduction redesign mid-lifespan (i.e., 3 years), with 10+ million units shipped over that lifespan. One design, one customer, 100x the volume, over 6 years. Simple microeconomics. And with a solid design in the Power architecture, with a sprawling set of options, it was hard to look beyond IBM, or at least Power itself.

Which is exactly what Microsoft, Nintendo and Sony did. They all use Power cores, Nintendo and Sony with specialized peripherals, and Microsoft with a more mainstream, tri-core design. IBM wasn't always involved with every process in each design, but they are reaping some major royalties in all of their designs -- direct and indirect.

This resulting economies-of-scale "cost swing" is two-fold.

Not only does IBM reach profitability, let alone avoid reoccurring design costs, very early due to volume, but the cost of the IC units themselves go down. They go way in fact, at 100x the volume, meaning the consumer can buy more power for less money -- especially at the console's initial release, when the "equivalent PC power" is far more costly. Most consumers are far more willing to pay $300-500 for something that is "current" for 3-6 years (even if that is only from a marketing aspect), than a PC that is only 1.5-3 years.

The same argument also holds for the Graphics Processor Unit (GPU) and, thus, companies like ATI (now AMD) and nVidia, as much as it does for the CPU with IBM and other, select licensees of Power involved with Microsoft, Nintendo or Sony. Instead of having to rev new PCI-Express (PCIe, or even the lesser known/implemented HyperTransport eXtension, HTX, for AMD) GPUs every 12-18 months for performance, a product can be rev'd for 3-6 years, with only a die shrink and reduced cost mid-life.

In fact, one could argue that this is merely the commodization of the console 15 some years after the same happened for the PC OEM at the start of the '90s. And it's just beginning.

Software entry and development costs v. volume and support costs

But hardware alone isn't the major cost to the gaming software itself. Software development houses have their own realities and resulting issues. But their metrics are little removed from hardware -- entry and development costs v. volume are the staples, with support costs replacing redesign costs.

To start, let's remove the similarities. Most game development is done the same, regardless of the target platform -- PC, console, etc... You have PCs, some mid-range, many high-end, possibly clustered, and with 3D absolute, definitely either rendered or graphics workstation-class for object modeling. Another aspect that is common to any target is post-release support -- i.e., patching -- especially with consoles now being Internet connected. In fact, now that consoles are Internet connected, games don't have to be completely bug-free before release -- something game development houses won't admit used to differ. With the complexity of console gaming (see the following section) rivaling PC games -- and often the same game ships on both, at least when the game/series began on the PC -- development houses can assume users will now download updates on their console as standard procedure, much like they do for PCs.

Now that's where the direct similarities end. More indirect, there are a few, higher initial costs for the console development kits, although depending on the game, it's not uncommon to have matching (or even higher) kit costs for PC targets, although there is clearly an equilibrium forming there too as most PC gaming engines are now available on the console out of sheer market desire. So even the "development kit/licensing" costs are indirectly relative these days. Furthermore, unlike hardware, there is also an indirect similarity in lifetime, as games are only good for 18-36 months on the console before they are "bargain bin" much like the more popular PC games as well. So no major difference there.

Which brings us back to the two major differentiators, volume and support costs.

A console title often sells in the hundreds of thousands, even for a lesser title, and millions are not uncommon for the top few dozen games on a console, even still holding some value some 2-3 years after release. PC games, on the other hand, are awarded "Game of the Year" titles when they cross hundreds of thousands of sales, and only where games are clearly not favored on consoles (which is becoming less and less of a difference these days), millions is very rare. In fact, a good testament to a best selling PC game that people assume only could be on a PC is when a console port is still made later, and still somehow sells well enough to make the port profitable (although not always).

But even volume alone is not the major issue for game development houses. Volume combined with support costs are.

Support costs are measured beyond just direct user support, which is virtually non-existent today. Commercial software support is abysmal even in the commercial/business application world where you sell at 10x the price and fee-per-incident or support contracts are considered "added cost." So when it comes to games, virtually no software development houses offer much other than forums or other places where users can help themselves, along with the occasional support moderator. No, the support costs are at every aspect of the game's history -- from development and release target to supporting the various iterations and variations of hardware in the PC platform. Add one part predicting the future with one part combinational probability and you've got a really, really unsolvable problem.

It starts in development by trying to predict the PC that will exist to run a title some 1-2 years before release. While some would say that's easier to do for a PC than a console, since a console is fixed and a PC can be upgraded, it's still the key issue for those "top 10 titles" that will sell hundreds of thousands or even break a million copies. Those titles will target the PC consumers will have at the mid-to-higher end (those willing to pay for the hardware and your software) and they will expect it to be at least usable for that generation at actual release, in addition to "look even better" when they upgrade in 18+ months. In reality, this double-set of requirement is much, much more difficult than targeting a "known quantity" of a fixed console platform, even when the platform is just specs some 18 months before it's actually available. It's still much easier from the standpoint of a title's development as it exits initial design and enters actual coding.

The support issue is then compounded by the myriad of PC configurations -- beyond just the CPU and GPU, but the countless factors, as well as the "general use" of the PC that tends to bring in spyware and various other, unexplained and unquantified factors. Yes, consoles and other embedded systems are not immune to such issues either, especially now that they are Internet-connected. But it's much, much easier to bring a console -- or any other embedded system for that matter -- back to a "known state," especially when the hardware is the exact same across millions of units. In contrast, a PC may be unique down to thousands or even hundreds of units in just end hardware configuration -- especially when it comes to targeting those "higher-end" consumers.

No, the major game development houses don't want to deal with a PC if they don't have to. Sure, there are many, many "name brand" PC gaming genres and series that are still selling, and will still sell, on the PC in the future. But these are clearly becoming the mainstay of only those established, well-known and return customer PC gaming genres and their select series (or even those newer titles that clearly rip off ideas from those series). More and more newer gamers are console gamers -- 90% according to many studies, as well as the fact that consoles are now outstriping PC sales in general (and not merely just PC gaming rigs) in the past year.

The reason? The "support" issue is more than just a developer problem ...

Consumer preferences of "load and forget" and "reduced limitations"

The greatest, most well-known -- to the general consumer -- system that runs the allegedly "hard-to-use OS" known as Linux is the common Digital Video Recorder (DVR), even more commonly referred by its most popular brand name, "TiVo." Virtually all major DVR solutions today run Linux, as do countless other embedded devices. Despite other attempts, Microsoft's Media Edition Windows OSes have utterly failed to "catch on" other than being "just another edition" of their general purpose OS (as well as direct, Linux-based open source toolkits other than for fringe, technical users like myself), because 90%+ of consumers just don't want to deal with a PC.

One basic hardware configuration. One "known quantity" software installation. No need to have a technically proficient kid, neighbor's kid or at least yourself to deal with it when it becomes unusable. Consoles, like set-top boxes, are what 90%+ of consumers want. "Load and forget."

It used to be that consumers had to accept select limitations with consoles over PCs. No 3D, typical side-scroller or other, simple gaming titles, countless other limitations including -- the biggy -- lack of resolution. Ugly, 640x480 interlaced -- one 640x240 frame one 1/60th of a second, another 640x240 frame another 1/60th of a second -- equivalent to a sight-disrupting 30 frames per second (fps). Even if the television was capable of providing 480p (progressive, 60fps) for a 480i input signal -- i.e., the game is still rendered interlaced, but the TV display avoids the eye-strain -- it's still not the same.

But that's no longer the case anymore.

I've already talked about the fact that sheer volume is now giving consoles the ultimate "price-for-performance" advantage at least at console release, if not even a year later, versus PCs. A consumer can also get a powerful gaming console for half (or even one-third) the price of a PC at release, especially since the vendor is selling below cost and will recoup later as hardware costs come down (or just via software, if they never do on hardware).

But the other flip of the coin, now in 2006+, is the fact that 720p -- that's 1280 x 720 resolution of progressive scan or (close to) 60fps -- is affordable. As such, most Americans, Europeans, Japanese, Chinese, etc... consumers who game, would buy a higher-end PC for gaming, etc... are putting in HDTV, commonly 720p or close, and have the capability for high-definition now. And with even yesterday's (i.e., 2005) ATI R400 and nVidia NV40 series GPUs, standard full screen anti-aliasing (FSAA) and other effects (such as shading) really makes higher resolutions less of a desire. That's not to say you won't have a significant number of gamers jump to 1080p (1920 x 1080 progressive), or even just 1080i (1920x1080 interlaced, 30fps), but 90%+ of consumers will see little difference, or at least not see the difference to pay for it for now.

720p gaming is what people can get for under-$500 in a console, and another sub-$500 in a 27", sometimes even 32" after rebate, LCD TV that also seconds as their main viewing device. In fact, with so many gamers either being singles, couples without children or children with their own "second TV" from the parents (not that I believe it is so wise), this sub-$1,000 configuration is quickly becoming "standard issue" more than PCs in homes. Indeed, one could argue that the console's quick volume increase in 2006+ has much to do with the fact that HDTV has become affordable, all while the PC need be little more than $400 for e-mail and web surfing, and no longer gaming.

Let alone do the games need to be taking up disk space and possibly "mucking up" the PC -- daddy's (or mommy's) PC for that matter, especially when you want to keep your teenagers off of it. Heck, even I -- an absolute Linux user for the past decade -- don't want to reboot into Windows just to game. When I can't find a title for Linux (which is actually the best platform for gaming, especially when you use the PC for work -- e.g., you don't have to be "root" to run a game, among other security advantages), I'd rather just have a console -- regardless of the underlying OS (even Windows-based) -- if it keeps me from having to deal with Windows on any PC because of the support issues (and not out of any dislike or political-based distain of anything Microsoft).

Enough pro-console? What do I think PC games offer? Does it matter?

After reading all this, you might think I'm pro-console, you just have a hard time accepting much of what I say is reality or you really differ with me because of some points you think I didn't cover. Trust me, I don't like it any more than you do, and I do think there are a few points left in the PC game industry that will keep it running (as I'll cover in the last part of this series). But that doesn't escape the reality that ...

Most major, commercial game companies don't care about us anymore, becauase we're not volume.

Of course, there will always be select houses with select games that cater to us. The big one is any game that requires the precision of a mouse. As much as Halo might have made first person shooters "playable" on a gamepad, it still sucks in my book (yes, I've played and finished it too) compared to the first person shooters we've all been accustomed to playing with a mouse on the PC. And that would include games you could enter and leave vehicles and what-not, PC games that pre-date Halo on the Xbox. I would argue it's popularity is really the result of people who never experienced PC gaming, and grew up with consoles.

With that said, Microsoft basically guaranteed that PC gaming would continue with select, major development houses for select, major product titles/series/genres for at least a short while longer -- purposely perhaps? (given PC gaming is a staple of the Windows OS versus other options, and not because of capability differences, but existing consumer volume) -- since their #1 mistake with the Xbox 360 (other than not offering digital video output options, only digital audio), was not offering a mouse input option. The only 3rd party options available poorly implement emulation for mouse input. And if you think that's bad, it gets worse for us old PC gamers ... consoles are purposely crippling games even on the PC!

One of my favorite games available for both the Xbox and PC (and I own both) was Elder Scrolls III: Morrowind. In Morrowind -- first/third-person Role Playing Game (RPG) -- where you strike someone with an arrow, a sword slash, etc... made a difference in how they were wounded, possibly fatally (especially when instantly). Showing the clear bias towards console players, and removing the expectations of the typical PC player, this was regressed out of Elder Scrolls IV: Oblivion, crippling "any hit makes the same difference." This basically reversed the entire evolution of first person shooters in the PC world -- localizing damage -- and it's not the sole example. But as I said before ...

Most major, commercial game companies don't care about us anymore, becauase we're not volume.

Just look at the popularity of Halo. 90%+ of console gamers -- which means 90%+ of gaming consumers -- don't care because they don't know. They aren't familiar with the Quake and latter engines that actually localized damaged and all those other small, subtle, but "world of difference" changes in almost 15 years of evolution in of the first-person shooter. That lesson is utterly lost, as 90%+ of console gamers -- now the majority far over PC gamers -- don't care, so, again ...

The major, commercial game houses don't care about PC gamers.


This is the reality people. I don't know if Microsoft's leaving the mouse out of their console was intention or not (because of their ties to the PC gaming platform with Windows). But since 90%+ of consumers don't care, and are willing to "play like Halo" essentially "forced them" to -- all without their knowing, and largely because of it -- it really didn't matter if they did. And, sadly enough, I have a Xbox 360 and even I have to agree that I don't miss the mouse for what I do play on the console (which is not first person shooters, of course, that's why I still have a PC).

What about Flight Sims and Space Sims and other things? Oh, perfect examples of why PC gaming is very much dying at age 25, as I will explore in the coming parts. Perfect examples they are, even though some of you reading this are just like me -- you "grew up" on the PC with flight sims and space sims and you loved games like the Jane's series to Wing Commander and latter incarnations like LO-MAC or iL-2 or Freespace and the X series. -- we're old dogs, and our genres' followings are not in those 90%. I.e. ... (my now broken record) ...

The major, commercial game houses don't care about PC gamers, especially our clearly PC-only genres with their "advantages."

Remember this reality as I revisit the history of PC gaming and its various genres in the blog posts to be continued here shortly ... ;-)

Additional thanx to Brian Ashe for feedback.

3 comments:

kahn said...

Yeah nice post, perhaps in the future we see everyone using apples, and for gaming people going with consoles as the preferred choice. Yet PCs are always ahead in graphics and processing power.

My PC site

wow power leveling said...

Americans everywhere humor A detention wow gold notice was written like this: a wow power leveling police car with stones, to win wow gold the detention center for seven wow power leveling days all-inclusive accommodation replica rolex Tour Value; hit send 2 a beautiful bracelet, wow power level fashionsuit, police transport; more more surprises , the former can enjoy free shaved 10; before the 100 can play with power leveling the dogs, the guests were presented massage sticks, electric shocks to CHEAPEST power leveling the dead skin beauty care services.

Adi said...

Oes Tsetnoc one of the ways in which we can learn seo besides Mengembalikan Jati Diri Bangsa. By participating in the Oes Tsetnoc or Mengembalikan Jati Diri Bangsa we can improve our seo skills. To find more information about Oest Tsetnoc please visit my Oes Tsetnoc pages. And to find more information about Mengembalikan Jati Diri Bangsa please visit my Mengembalikan Jati Diri Bangsa pages. Thank you So much.
Oes Tsetnoc | Semangat Mengembalikan Jati Diri Bangsa