Hypercritical
May 2013
Next Generation
Now that the Xbox One has been revealed, joining the already-released Wii U and the previously announced PlayStation 4, we can finally get a sense of what the next generation of game consoles will look like.
This used to be a simple business. Cutthroat and fiercely competitive, yes, but at least all the players were racing for the same prize. Every handful of years, we’d get a new crop of consoles, each claiming to be the most powerful and to have the best games.
Seven years ago, after being outsold by Sony in the two previous console generations, Nintendo broke from the pack and went after a new market: people who were not interested in—or were too intimidated by—traditional game consoles.
The Wii was startlingly less powerful than the other consoles in its generation. This helped make it the least expensive and the smallest, which only increased its appeal to non-gamers. The coup de grâce was the Wii’s novel control scheme, which let your dad, who couldn’t get past World 1-1 back in the 80s, make an improbable transformation into a hardcore gamer…of a sort.
And if the idea of “winning” a console generation with laughably underpowered hardware wasn’t enough, the Wii and its contemporaries also put an end to the idea of a game console that just plays games. Just a few years after launch, all of the consoles—even the dainty, standard-definition Wii—supported some kind of social networking, photo viewing, and one or more video streaming services.
Arguably, this movement started to gain momentum with the original PlayStation’s ability to play music CDs, and continued with the PlayStation 2’s secondary role as a DVD player. But the Wii, PS3, and Xbox 360 definitively moved the entire product category beyond gaming. In fact, the PlayStation 3 ended up as the most popular way to view Netflix on a TV.
This was all a natural consequence of the decreased cost of storage and computation combined with the ubiquity of wireless networking. It was inevitable that any TV-connected box would eventually support these features. But it also means the Xbox One, PlayStation 4, and Wii U lack the clarity of purpose enjoyed by the previous generations of game consoles. Here’s how things look to me at the dawn of the next generation.
Wii U
Stop me if you’ve heard this one before. The Wii U is dramatically less powerful than the Xbox One and PlayStation 4. In place of hardware power, Nintendo is offering an unconventional multi-screen gaming experience using a tablet-style controller. Although pricing has not been announced for its competitors, there’s a reasonable chance the Wii U will end up being the least expensive console in this generation.
It sure looks like the Wii formula all over again, but there’s a difference this time. The Wii U’s GamePad controller is significantly more intimidating to non-gamers than the familiar-looking Wii remote. Wii accessories (and games) also work with the Wii U, which is nice, but the GamePad is the face of the new system to consumers. For former Wii buyers who are intimidated by the GamePad, Wii hardware and software compatibility may only make them further question what the new system really offers beyond the Wii. And though the Wii U expands on the Wii’s non-gaming features, its TV integration feels half-hearted and has thus far failed to impress.
The end result has been dismal Wii U sales coming out of the 2012 holiday season. Nintendo’s rumored consideration of allowing smartphone apps to run on the Wii U seems uncharacteristically desperate.
Thanks to the novelty and accessibility of the Wii remote and the universal appeal of launch titles like Wii Sports, the Wii sold in such huge numbers that third-party developers couldn’t afford to ignore it. They dutifully cut down the features and graphics quality of their most popular games to get them to run on the Wii. These games were often terrible, but at least they existed, giving the Wii’s game library “checkbox parity” with the rest of the market.
Like the Wii, the Wii U is not powerful enough to run the same games as its competitors. Unlike the Wii, the Wii U’s sales numbers aren’t high enough to motivate cut-down ports of new games. That leaves the Wii U with Nintendo’s franchise titles (many of which are not yet available), a scant few Wii U exclusives from third-party developers, and several ports of previous-generation games that Nintendo’s new hardware is finally able to run.
It’s still too early to call this race, but the Wii U certainly looks like it’s in trouble. It may be that Nintendo has just built the wrong machine. For the most part, the Wii succeeded despite its underpowered hardware, not because of it. Choosing to produce another “next-generation” console with previous-generation power isolates Nintendo.
New multi-platform titles can easily target the Xbox One, the PlayStation 4, and the PC simultaneously. The Wii U isn’t even in the running—unless it sells so well that a hobbled port is justified. The same goes for exclusives built around the Wii U’s unique features. No third-party developer wants to invest in a game that can only ever be sold on a single platform with a tiny installed base.
I own a Wii U, and I’m convinced that it really does offer new, fun gaming experiences not available on any other platform. I’m also a diehard fan of several of Nintendo’s popular franchises. But I’m not the kind of customer that carried the Wii to head of the class in the previous generation. I’m the kind that would gladly pay twice the price of a Wii U for the ability to play a Zelda game on a console with the power of the PlayStation 4. The Wii U is not built for me. Whatever kind of customer it is built for, there sure don’t seem to be many of them.
PlayStation 4
Sony is the reigning king of overblown hardware hype, famously promising that the PS2’s emotion engine and the PS3’s Cell processor would change the face of computing forever. And maybe they did, in a tiny way. But their power was notoriously difficult to unlock. They became the standard-bearers for the gaming version of the ancient Chinese proverb: “May you develop for interesting hardware.”
Hardware eccentricity has been part and parcel of console development for decades. And the weirder the hardware, the more likely it is that a straightforward implementation of a game engine will run up against bottlenecks. The developer laments are familiar. “If only there were more bandwidth between the CPU and main memory.” “If only I had just 10% more RAM.” “If only this console had a much more powerful programmable GPU instead of a ring bus studded with custom SIMD processors, each with its own tiny local storage.”
The PlayStation 4 aims to repent for the sins of both its father and grandfather—and then some. Unlike its predecessors, it was designed in close cooperation with game developers. During the design process, new revisions of the PS4 architecture were presented to developers along with a challenge: find the bottleneck. Every aspect of the system was put through a similar gauntlet, from the shape and travel of the controller triggers to the accuracy of the gyroscopes.
All game consoles go through some version of this process, but the PlayStation 4 is defined by it. The hubris of the PS2 and PS3 is nowhere to be found in the PS4. This is a product of a newly humbled and rededicated Sony.
And the thing that Sony is rededicated to is gaming, plain and simple. Sony was the first console maker to really push the idea of a gaming system that does much more than just play games, but now it’s returning to its roots.
The PlayStation 4 is exactly the sort of thing that a hardcore gamer might have envisioned if presented with the product name back in the days when the original PlayStation reigned supreme. It’s got more of everything, and the vast majority of its resources are bent towards being the best system for developing and playing games. In this generation of consoles, that’s actually a radical notion.
Xbox One
The final entrant in this round of the console wars is the most ambitious. No longer content to walk the old paths blazed by Nintendo, Sega, and Sony, Microsoft is finally making its play for the entire living room.
Take a peek at the back of the box—a box that looks for all the world like a futuristic VCR—and you’ll find the hardware incarnation of this ambition: an HDMI input. Any form of entertainment that does not spring from the Xbox One is invited to at least flow through it, to be mediated and controlled by it. It’s all right there in the name: One box to rule them all.
The Xbox One announcement was unabashedly focused on everything but games. Microsoft promised more at E3, relying on the substantial goodwill it’s earned with gamers over the past decade to stave off any anxiety about the One’s gaming bona fides.
Indeed, at first glance, the core hardware architecture looks nearly identical to the PS4. But a closer look reveals a system designed to accommodate a much broader vision of home entertainment.
Where the PS4 uses high-speed GDDR5 RAM, the Xbox One opts for slower—but also less power-hungry—DDR3. And in the Xbox, that RAM is shared between two separate operating systems running simultaneously: one for games, and one for everything else.
These hardware features express two very different usage models. The PS4 expects to be turned on when in use, then turned “off” afterwards, entering a super-low-power mode during which a tiny auxiliary processor handles housecleaning chores like downloading game content and applying software updates.
The Xbox One, with its HDMI input and non-game-related OS and apps, expects to be fully powered whenever the television is on. Thus, Microsoft’s focus on idle power consumption—even at the cost of gaming performance.
To mitigate this disadvantage, the Xbox One includes 32MB of low-latency embedded SRAM right on the SoC. This is a common technique, but it leads to increased complexity. Game developers must now take care to ensure that the right data is in the tiny local eSRAM pool exactly when it’s needed. A single pool of uniformly fast memory (albeit with higher latency), as in the PS4, is a much simpler arrangement. Different priorities, different trade-offs.
(The eSRAM also consumes die space, which, along with power consumption and cost, may have contributed to Microsoft's decision to give the Xbox One 33% fewer GPU cores than the PS4.)
Then there’s the Xbox One’s companion hardware, the next iteration of Microsoft’s Kinect motion control system. The first version of this technology, released as an add-on for the Xbox 360, was the proverbial dancing bear: it didn’t work well, but it was amazing that it worked at all.
The new incarnation comes bundled with every Xbox One, and it dances like a furry Fred Astaire. It surpasses its predecessor by many multiples in every specification: resolution, depth perception, motion tracking, latency, noise cancellation, local computation. This technology is no joke.
But does it make games more fun? Or, failing that, is it a better way to control a television than a remote control? Microsoft is betting a lot, in terms of both hardware cost and software support, that the new Kinect will be an essential component of at least one of these activities in a way that the first Kinect was not.
When I’m feeling optimistic about the Kinect, I think back to the many generations of terrible touch-screen devices that preceded the iPhone. The history of touch-based interfaces on consumer electronics wasn’t a gradual ramp up to acceptable quality. The iPhone wasn’t just the next iteration; it was a discontinuity. Once the technology passed some critical threshold of responsiveness and reliability, it went from a nerdy curiosity to completely mainstream in the blink of an eye.
I don’t know where that threshold is for multi-sensor full-body motion control and voice recognition, but I do believe it’s out there. Microsoft does too. Of course, that belief will be of little consolation to Xbox One owners if the “iPhone moment” is still many years in the future.
Forward-Looking Statements
Last generation, Nintendo did something crazy—and it worked. This generation, everyone is taking big risks.
Nintendo tried to play the same hand that it won with in the last round, but now finds itself stranded with previous-generation hardware in a next-generation market. Like Apple in the 90s, Nintendo is a sentimental favorite. But it took more than just the iMac and the iPod to transform Apple. The Wii U still has the potential to be an excellent platform for Nintendo’s beloved first-party games, and a low-cost alternative to the PS4 and Xbox One. Nintendo should milk it for all it’s worth, and get busy on the next great thing.
Sony is betting that the market for game consoles made by and for hardcore gamers has not yet peaked. If it’s right, Sony is well-positioned to dominate this generation. If it’s wrong, the PS4 could be Sony’s Spruce Goose: the ne plus ultra of game consoles, remembered in equal parts as a technical marvel and a cautionary tale.
Finally, there’s Microsoft, offering us a brief glimpse of the boundless hunger that once defined the company. But as Microsoft knows all too well, the living room is littered with the bones of past suitors.
I applaud the technical prowess of the Xbox One’s software, particularly the focus on responsiveness. The demonstrated performance when switching between live TV, gaming, and other apps puts all previous efforts at “smart” TV interfaces to shame.
That said, I seriously question the public’s appetite for displaying any additional content alongside a TV show or movie. The “second screen” experience is already well established, and it happens with a device that’s in your hand or on your lap. Grabbing one third of a large, communal TV screen to look up an actor on IMDB isn’t just unappealing and cumbersome, it’s downright rude.
There are other contexts where the Xbox One’s unique abilities might shine: jumping in and out of a game to check a sports score, for example, or quickly hitting the web to watch an extended version of an interview after finishing an episode of The Daily Show. Yes, I can see that.
But will it be enough to crown the Xbox One the king of the living room? As with all TV-connected devices, content is the key. The Xbox One has games, live TV, and video streaming services covered, but it appears to lack any form of time-shifting functionality. Given how much popular content remains locked up in broadcast and cable TV packages, there’s no way any box without DVR-like functionality can ever be the One True Interface to “watching television.”
Luckily for all three companies, things change quickly in this industry. If a critical mass of programming becomes available on streaming services a few years down the road, the Xbox One could finally fulfill its destiny.
On the other hand, Microsoft’s new focus could be a giant turn-off to gamers who were expecting an “Xbox 720,” not a Kinect-powered “media center.” However brief and anecdotal it may be, a Wii U sales spike accompanying the Xbox One announcement has to have Microsoft at least a bit worried. If the gamers who bought the Xbox 360 don’t show up in the expected numbers to buy the Xbox One, I have a hard time believing this monstrous, sensor-festooned device will pull a Wii and capture the imaginations—and dollars—of non-gamers on a grand scale.
No matter what happens, I don't envision a future where the market is evenly divided between these three very different products. Game on.
If you’d like to hear an expanded audio discussion of these topics, including my take on the TV-related efforts of Apple and Google, check out episode 3 of the Ad Hoc podcast with Guy English and Rene Ritchie.
Hypercritical T-Shirts
Update - May 14, 2013: I regret to report that the Hypercritical t-shirt has been canceled due to the unauthorized use of an icon from a past version of the Macintosh operating system. All purchases will be refunded in full. This situation is entirely my fault. I'm sorry for disappointing everyone. Thanks to all of you for your support.
(My original post about t-shirts appears below, for historical purposes.)
I’ve wanted to create a Hypercritical t-shirt for a while now. When I saw that my friend and podcast co-host Marco Arment had created a t-shirt for Marco.org using a new Kickstarter-like website called Teespring, I was intrigued. When I looked at the shipping dates for Marco’s shirt, I realized that it was now or never if I wanted to get Hypercritical shirts into people’s hands in time for WWDC.
The Teespring website made it incredibly easy to get a shirt up for sale. In hindsight, doing this in the middle of the night on a Friday was perhaps not the best idea, but that’s what I did. In less than 30 minutes, I’d created the artwork, uploaded it, and started the sale. The rest of the weekend was a bit of a blur. I worked with the Teespring staff (and people on Twitter) to improve the artwork I’d created in haste, and to make more colors and styles available.
Teespring is a relatively young company, and the user-facing interface on the site doesn’t yet support adding multiple colors and styles. The Teespring staff made all these changes for me behind the scenes—on a weekend. The site’s limitations still necessitated the creation of two separate t-shirt “campaigns” for the two different ink colors used on the dark and light shirts.
If you follow @hypercritical or @siracusa on Twitter and were online this weekend, this is all probably old news to you. My dual t-shirt sales have already far surpassed their goals, thanks to the amazing response of my Twitter followers.
If you’d like to support my writing on this site, the t-shirt sale will continue until May 14th. According to Teespring's shipping estimates, all orders should arrive by June 4th at the latest, including international orders. Here are the two links for the dark and light t-shirt sales:
- Hypercritical: The Shirt - Silver and Gold
- Hypercritical: The Shirt (Dark) - Navy and Black
My sincere thanks to everyone who has already purchased a shirt, and to all the people who continue to read this site.
Beauty, Truth, and Jony Ive
The prevailing wisdom about software design at Apple is that the pendulum has swung too far in the direction of simulated real-world materials, slavish imitation of physical devices, and other skeuomorphic design elements, producing a recent crop of applications that suffer from an uncomfortable tension between the visual design of the software and its usability and features. After the executive reshuffle six months ago, we Apple fans have been hoping that Jony Ive, now in charge of Human Interface for both hardware and software, will end this destructive conflict and bring order to the galaxy.
With iOS 7 and OS X 10.9 looming, we’re left to wonder exactly what kind of software designer Ive will turn out to be. Certainly, Apple’s software has been influenced by Ive’s hardware designs in the past—and perhaps vice versa—but this will be the first time Ive is officially in charge of the virtual bits as well as the physical ones.
We may not have much to go on when predicting Ive’s software tastes, but we do know a heck of a lot about his opinions on hardware design. Though Ive has historically spent his time at Apple keynotes in the audience rather than on the stage, he’s starred in many, many videos wherein he explains why Apple’s great new hardware product looks and works the way it does. In these videos, his message has been remarkably consistent.
Ive demands that the hardware be true to itself—its purpose, its materials, the way it looks, and the way it feels. Here’s a quote from one of Ive’s rare appearances outside an Apple press event, talking about hardware design at Apple.
When we’re designing a product, we have to look to different attributes of the product. Some of those attributes will be the materials that it’s made from and the form that’s connected to those materials. So for example, with the first iMac that we made, the primary component of that was the cathode ray tube, which was spherical. We would have an entirely different approach to designing something like that than the current iMac, which is a very thin, flat-panel display. […]
A lot of what we seem to be doing in a product like [the iPhone] is actually getting design out of the way. And I think when forms develop with that sort of reason, and they’re not just arbitrary shapes, it feels almost inevitable. It feels almost undesigned. It feels almost like, well, of course it’s that way. You know, why wouldn’t it be any other way?
Steve Jobs also subscribed to this philosophy. Witness his explanation of the design of the first iMac with an LCD display at Macworld New York in 2002. Here’s how Jobs described Apple’s solution to the inherent compromises (in 2002 technology) of putting an optical drive in a vertical orientation and trying to pack an entire computer behind an LCD display.
The big ideas was, that rather than glom these things all together and ruin them all—a lower-performance computer and a flat screen that isn’t flat anymore—why don’t we let each element be true to itself? If the screen is flat, let it be flat. If the computer wants to be horizontal, let it be horizontal.
It’s interesting that Jobs and Ive saw eye to eye on hardware design and yet seemed far apart, at least in Jobs’s final years, when it comes to software design. While Jobs was reportedly a champion of rich Corinthian leather, Ive could only wince when asked about it in an interview.
I’m confident that we’ll see less leather, wood, felt, and animated reel-to-reel tapes in Apple’s future software products, but the question remains: what does it mean for an application or an OS to be true to itself?
I’m not sure how Ive will express that concept, but Loren Brichter, creator of Tweetie and Letterpress, offers one possible interpretation on an episode of the Debug podcast (starting at 6:10, and again at 1:02:26, specifically mentioning Ive). Letterpress is an exemplar of the so-called “flat design” aesthetic (and it’s also currently featured on the front page of Apple.com). Brichter designed the look and feel of Letterpress based on the things that modern graphics hardware is naturally good at doing: drawing and manipulating flat planes of mostly solid colors.
A design philosophy so tightly linked to nitty-gritty details of silicon chips and OpenGL APIs is unlikely to resonate with Ive as much as it does with a programmer like Brichter, but the end results may be similar. I expect Ive to focus on harmony between the look and feel of the software, the materials and finish of the hardware, and most importantly, the intended purpose of each specific application. (It’s kind of a shame that Apple’s already used the “Harmony” code name.) This is my message to Jony Ive and and my hope for iOS 7, OS X 10.9, and each bundled application: to thine own self be true.