Page 2:A Lesson In History: The Death Of The Sound Card
Page 3:Have 3D Graphics Reached The Point Of Diminishing Returns?
Page 4:Can Software Developers Keep Up?
Page 5:AMD In The Next Decade
Page 6:What About Intel?
Page 7:Intel Graphics In The Next Decade
Page 8:Nvidia's Ambition
Page 9:Nvidia As x86 Manufacturer
Can Software Developers Keep Up?
What follows is a very long-winded way of saying: the video game market is large enough and profitable enough to support the development of titles with Pixar-level budgets of $150-175M. Seriously. It’ll be very long-winded.
In order to keep up with hardware, software developers need huge development budgets. Today’s "Quadruple A" games, such as Call of Duty: Modern Warfare 2, Halo 3, and Gran Turismo 5 have stratospheric development budgets of $50 to $60 million. Grand Theft Auto 4 has even been reported as having a development budget of a $100 million. With marketing costs factored in, CoD: MW2's budget is a reported $200 million!
Ambitious "triple A" games built to establish a new franchise, such as Bioshock or Mirror's Edge, range in the $20 to $30 million budget. These games have the potential to grow into major new franchises, and their budget reflects that.
For 2K Games, it was a great investment. But for EA, its investment resulted in a critically-acclaimed game with a wonderful visual design and concept, but Mirror's Edge was a financial disappointment. If you haven't played the game, you really owe it to yourself to try it at today’s bargain-bin price.
Historically, software development budgets have increased by an order of magnitude with each console generation. In the Super Nintendo/Genesis era, a big budget game was in the range of $300,000. By the GameCube/PS2 era, games were in the $10 million range with top-tier games like Final Fantasy warranting a $40 million budget. If this trend continues, we will be seeing a $600 million budget for Gran Turismo 7 on the PlayStation 4, a $6 billion budget for Gran Turismo 9 on the PlayStation 5, and a $60 billion budget for Gran Turismo X. Obviously, this trend cannot be extrapolated forever.
The exponential growth of software development budgets over the last two decades has indeed been a reflection of increased hardware performance. But it was also reflective of the expanding market size and the resulting increase in potential profits. The financier of a game isn’t budgeting the game "so that it looks cool." Rather, they’re considering the risk/reward of the title, and they are doing that across an entire portfolio of software titles. For every Guitar Hero or Madden NFL, there will be a Mirror’s Edge or Duke Nukem Forever. The sequels that reviewers like to criticize for being "more of the same" subsidize the budget for more ambitious and creative games.
The process of developing and selling games is similar to that of developing a Hollywood blockbuster. Although games cost $50-60 a piece and movie tickets are around $10, a household can buy one copy of the game and enjoy it, while each person in the theater needs to pay for a movie ticket.
As a general rule, successful effects-driven Hollywood movies with budgets in the $150-175 million range bring in approximately 4x to 5x their budget in worldwide ticket sales. These are movies like Transformers, Casino Royale, and Pirates of the Caribbean. Successful lower-budget story-driven films can have a widely varying rate of return, starting from 3x for The Departed, 6x for 500 Days of Summer, 10x for Twilight, and 25x for Slumdog Millionaire. Movies like Children of Men can end up making less in ticket sales than their budget (10% deficit).
|Title||Estimated Budget (source: IMDbPro)||Worldwide Ticket Sales (source: IMDbPro)|
|Pirates of the Caribbean: Curse of the Black Pearl|
|Pirates of the Caribbean: Dead Man's Chest|
|Pirates of the Caribbean: At World's End|
|Star Trek (2009)|
|(500) Days of Summer|
|Children of Men|
You cannot use these numbers to talk about the actual profits that a company brings in, due to things like DVD/Blu-ray sales, broadcast rights, and costs of marketing. But these numbers allow us to see that the risks a financier is willing to take are not shouldered arbitrarily. When committing nine-figures before the first ticket is sold, the financier has to be sure of what he is doing in the long run. Just because several movies have reached the $1B mark in ticket sales doesn’t mean that you’ll see a $500 million pre-marketing development budget anytime soon. Likewise, one may be more comfortable committing eight-figures toward a movie, knowing that one’s losses will be smaller when you lose, and gains will be bigger when you win.
Financiers of games follow similar patterns. If you look at successful "big budget" games like Bioshock, GTA IV, Halo 3, and Final Fantasy XII, games seem to be have a sales that reflect 4x to 6x their budget (ignoring marketing costs). Again, that 4x to 6x estimate doesn’t mean anything about what they really bring back, but instead reflects the fact that successful "big budget" games ranging from $20 million to $100 million, end up selling a similar number of copies that is proportional to the investment. The financiers of games try to follow a similar unwritten risk profile.
Industry pundits will tell you that the video game industry is comparable in total financial size to the motion picture industry, if not larger. That’s not entirely true. In 2008, Media Control GfK International reported that global sales of video games reached $32 billion--more than the $29.8 billion amassed in DVD/Blu-ray sales.
The catch is that box office ticket sales accounted for another $28.1 billion. We haven’t even begun to touch on broadcast rights.
Moreover, the success of a Hollywood film is more predictable than that of video games due to all of those multiple revenue streams and a better understanding of the market. So, while game development budgets will continue to grow, it’s hard to imagine games reaching the same budget levels as Hollywood films. The risks are higher in game development. And therefore, for any given budget, they’ll want a higher return. Over the next ten years, I anticipate quadruple-A games to have budgets in the region of $175 million, and ambitious games introducing new intellectual property having budgets of $60 to $80 million.
With Pixar-level budgets come the potential for Pixar-level graphics (and Pixar-level characters and stories). Given that Pixar films still require 5 to 6 hours to render a single frame on large supercomputer clusters, the answer is no, graphics have not reached the point of diminishing returns yet.
This means that we haven’t reached the plateau in "subjective experience" either. Newer and more powerful GPUs will continue to be produced as software titles with more complex graphics are created. Only when this plateau is reached will sales of dedicated graphics chips begin to decline. In addition, unlike the sound card world, in which the plateau was reached with a relatively small number of transistors, allowing rapid integration onto motherboards, GPUs are still pushing the limits of process technology and require massive cooling and power. That is, you could manufacture a CPU and GPU on a single physical package, but you’re going to have a heck of a time cooling that device and delivering the appropriate amount of power (Ed.: Intel's Clarkdale-based Core i5 and Core i3 are a perfect example of the limits of CPU/GPU integration today). We’ve also only looked at graphics as the driving force for development. GPUs can be used to accelerate large, complex parallel mathematics. This is the other frontier for manufacturers like AMD, Intel, and Nvidia.
The software world continues to provide ripe opportunities for this merry trio, and allows them to sustain substantial R&D efforts. The state of the graphics industry is strong.