This last week, something of a surprise emerged and had what was perhaps the quickest hate-to-love news cycle ever.
Valve unveiled the Steam Deck, a portable custom PC built on an updated version of the Debian-based SteamOS and a semi-custom AMD APU with 4 Zen 2 CPU cores and 8 RDNA2 GPU cores. Paired to 16GB of LPDDR5 system memory and one of 3 storage options, the system starts at a lean $399 for the 64 GB eMMC storage model and goes all the way up to $649 for a unit with 512 GB of NVMe storage. Rounding out the specs are a 1280×800 display, which renders at native 720p and has the remaining margins for other visual elements of the OS, the integrated controls which takes lessons from the ill-fated Steam Controller, using two touch panels but then also pairing them with twin analog sticks, a directional pad, and a robust mix of face, shoulder, and rear buttons, and then rounding out with a USB-C port that can be paired with an optional dock to make the unit a PC all unto itself, effectively.
So, what’s the big deal with this one?
In theory, and the first news cycle after its announcement, it was an object of derision. It combined Valve’s two worst hardware failures – the Steam Controller and the Steam Box – and it seemed like a non-started based on that. It was being panned for claiming “next-gen” hardware despite not having Zen 3 CPU cores, it was slammed for a limited 720p display, and the base eMMC storage model was dissed for being too slow to enable proper “next-generation” gaming as Valve was promising.
However, something funny then happened – on pre-order day, it wasn’t such a joke anymore, and Valve sold so many reservations on the unit that an order placed today shows delivery estimates in Q2 or Q3 of 2022. What’s the big deal, anyways? How did opinion change so rapidly on this little thing?
Portable Gaming Is Like, So Hot Right Now
The enduring appeal of the Nintendo Switch is an object lesson in this, as is the fact that mobile gaming remains the industry’s most profitable market, and it should have informed the flip in reaction here. Playing games on the go or in easy form factors is popular, and it appeals to a broad audience precisely because people don’t always want to be chained to a desk or couch at home for gaming. In such a form factor, some of the concerns of the system melt away – a 720p render target sounds bad in 2021 until you realize that at that screen size, no one save the snobbiest pixel snob is ever going to notice. Controlwise, sure, it has the Steam Controller touchpanels, but they’ve made smarter plays by including standard twin-stick controls and a standard d-pad to round things out. It’s a system made on compromises, but the compromises are intelligently-made for the form factor and price target in mind.
Next-Gen Hardware? By Most Definitions, Yes
An easy thing for hardware snobs, including myself, to dunk on about the announcement was calling the APU at the heart of the system “next-gen,” since AMD launched a better CPU core design…last year, and while RDNA2 is cutting-edge modern graphics hardware, it also is memory-bandwidth hungry, so a portable design doesn’t necessarily feel that great using standard LPDDR5 memory. However, at 720p, the bandwidth requirements aren’t nearly as high, so the whole reason desktop RDNA2 has Infinity Cache (making high resolutions work better) is sort of moot. On top of that, LPDDR5 at 5,500 MT/s is still plenty fast and should work very well at 720p while also offering low latency for the Zen 2 CPU cores, which need that lower latency to offer top speed. Lastly, while it’s true that AMD does have a newer CPU core design in mass production on desktop and mobile, what was easy to miss up-front if you were dunking on Zen 2 is that the pairing of Zen 2 and RDNA2 is well-proven as next gen, given that it is the foundation of both the PS5 and Xbox Series consoles. This actually makes it quite fascinating, as it offers what is a sort of further stripped down version of those systems, with half the CPU cores and drastically fewer GPU cores, but it is also pushing far less pixels than even the target resolution on Microsoft’s Xbox Series S machine. So it’s not the newest hardware bar-none, but it is a tried-and-true combination that delivers pretty well in the hardware such designs are already present in!
The PC Component Shortage Made Us All A Little Bit Crazy
I’m not going to suggest that a lot of people necessarily chose the Steam Deck over getting new PC components, or that a lack of GPU access drove people to a portable with a fairly slim amount of horsepower in that same department. There’s not enough actual data to back up any observation on that front. However, I do think it somewhat likely that the woes of availability drove some demand for the Steam Deck. Being based on a similar semi-custom APU but smaller could very well have driven a lot of people towards the Steam Deck instead of the hard-to-find PS5 and Xbox Series consoles. Likewise, if you want a new PC but are waiting for shortages to abate, getting a Steam Deck isn’t a completely illogical move, especially if you’re not trying to game on the cutting edge of technology. The Steam Deck offers powerful-enough hardware, a compelling and unique form-factor that has grown in popularity, and is built for the most popular PC gaming marketplace in the world. All of that makes it pretty damn good!
The Handheld PC Market Is Growing But Had No Marquee Product, Until Now
Over the last several years, a large number of handheld PCs have come out all touting various different features to draw people in. The biggest is the GPD Win series of mobile PCs, all of which have used mostly off-the-shelf laptop CPUs from Intel and delivered reasonable performance for some mobile gaming. There’s the Dragonbox Pyra, a Debian-based pocket PC that has only shipped to developers as of August 2020, and the ancient Pandora device, which last shipped in 2010 and the Pyra succeeded. Most of these projects are crowdfunded or prototypes in progress, which means buying one is sort of dicey – the GPD Win is now successful enough that you can buy one on Amazon, but most others remain in the shadows, either because they have crowdfund stink on them or they have big “knockoff electronics from no-name brand” energy. Many of these are also designed as very general pocket PCs with gaming capabilities kind of added in last minute – many of them have full QWERTY chiclet keyboards and then they put two short analog sticks and some face buttons for gaming into an awkward spot. The GPD Win devices have gotten good press coverage from tech channels on YouTube – they’re perpetually on LinusTechTips and LowSpecGamer, but the devices remain low-volume. Valve is smart to hit this market, because a gaming-first design with better specs and focus on gaming from a trusted brand was a value proposition that has not yet been offered in this space. With no crowdfund stink and active connections to developers and publishers around the world, this was bound to be a success.
Steam’s Game Library Is So Diverse That Specs May Not Even Matter
Focusing back on hardware as a pivot point to a new topic, let’s talk about the games that are actually on Steam. A lot of people, including the announcement, focused in on large-name AAA titles as the sizzle footage, and admittedly, seeing Control running on a $400 handheld when it can’t even run natively on the Switch was a strong play. However, what I think is most interesting is that the Steam Deck is a sort of strong way to play indie titles. Steam’s strength as a marketplace is that it has something for literally everyone, from visual novels to first-person shooters, from JRPGs to strategy games, and even includes a bafflingly large hentai game library. Steam has everything. For as many people got Steam early as the exclusive means of distribution for a handful of Valve and other AAA titles, many more came onboard as it started to offer indie gems and give those titles viable, highly visible means of distribution. The APU isn’t the absolute best or a mega-beast system, so what? I can play Hades at 720p in line at a con or on a bus. I can see how many people look in horror if I play the last part of the first part of Doki Doki Literature Club in public (although with DDLC+, you can now do that on Switch anyways!). Sure, it can play Control, and that likely means with settings tweaks, you could get Assassin’s Creed 87 or Call of Duty 24 running on it too, and that’s great – but so much of Steam’s library is fun, quirky indie titles, pixel art masterpieces, and all sorts of other titles that don’t need huge powerhouse machines but are instead tweaked and tuned to be accessible to the masses, often made by people without beastly home PCs. Leaning into that is an absolute strength, and while Nintendo sort-of does this with the Switch, that’s kind of a hidden-gem feature of the system. The Steam Deck, whether Valve intentionally markets it as such or not, is automatically the indie machine.
Open-ness, and Not In The Smell-Your-Own-Farts FOSS Way
Can I confess something? I love FOSS (free and open source software) but I hate it. I love it because I think it is a good model and a way to spread access to powerful tools to people, but I also hate it because it tends to attract a lot of fart-sniffing monologues about how great access is and how software can change the world, and like, sure, I agree to a point, but it tends to be far too self-aggrandizing and make it sound like so much more than it is. The Steam Deck is, in part, based on this, using Debian as the basis for SteamOS, and then using Proton libraries to bring Windows app compatibility to the device so they can run semi-natively through translation layers. This is great because Windows, even at its lightest, is still a bloated mess of an operating system. However, Valve is doing something sort of great with the device – it’s not locked to SteamOS at all, and you can choose to run Windows on it. You can choose to run competing game stores on it – Valve themselves brought up the Epic Game Store – and that would mean that you could run just about anything that it meets the requirements for. Want to run World of Warcraft on it? Sure, why not. Final Fantasy XIV? Hell yeah, get it – even if you don’t have the Steam version. For most of my audience, that actually makes it a rather interesting device, because it could very well be a fantastic way to play most MMOs – docked with keyboard and mouse, or in the wild for those with controller support. To that end, however – controller-supported MMO can be both setup and punchline for a joke, but even having the option to just play on the go is compelling. Being able to support full Windows means it could even be a cheap desktop – buying it and a dock still beats the price of a lot of lower-cost desktops and exceeds them in power through a strong APU and high-bandwidth memory!
Some Cause for Concern
The only thing that gives me pause is that Valve’s hardware track record is mixed, and largely negative. The Steam Controller was a cool idea but it was marred by poor execution, shoddy build quality, and it ended up being on-sale for cheap very, very quickly. It made a lot of assumptions about how players would interact with games via a controller while the market was already moving towards the console model, aided by the ease of getting an Xbox 360 controller working with a PC. The Steam Box idea was really quite neat, but it wasn’t a good standalone device and the idea of playing a PC game in the living room was sort of not popular – especially since it leaned heavily on streaming across your home network (the Steam Link version of the concept did, at least). Every manufacturer that came out with a Steam Box capable of full play locally sold an abysmally small number of units by most data points available, and the idea was pretty quickly disappeared after that bomb. Valve’s Index VR headset is perhaps their best hardware effort to-date, but even it is successful largely due to being a big fish in a small pond – VR as a market just isn’t that large at home for a myriad of reasons (space requirements, hardware installs like lighthouses and trackers, minimum system specs being much higher than average for just even a basic game) but that isn’t to say that the Index is bad – in fact, it is one of the best options for VR, especially now that Oculus is fully integrated with Facebook. You can see the problem simply here, though – out of 3 major hardware launches, only one of them is anywhere near a success, and even that is successful in the context of a small market with high-priced, luxury hardware that is out-of-reach for most average consumers.
The next big issue is on the software front. SteamOS is…fine, but it has some problems with modern games and for games with anti-cheat software implementations, since it runs on a streamlined version of WINE to get Windows API calls working on Linux. Proton helps with this a lot, but it has problems with a lot of big online titles. Further, because this is all done via abstraction layers, it creates performance degradation – you’re already starting with a cut-down APU and then you’re not going to be able to run the game natively at full speed on it because some of the horsepower is being diverted into running Proton and its components. To their credit, Valve is working on enhancements to Proton to ensure the software is both more broadly compatible and also faster specifically for the Deck, but there’s no guarantee of this and Valve is also leaning on developers to work with them to help on that front. If a game you like doesn’t work with Proton and the developer doesn’t help Valve develop better compatibility, it might just not work for that title. Luckily, you can switch and use Windows, but then you’re buying a Windows license for it and also forgoing any enhancements for other titles that come via the Proton layers on SteamOS. Lots of oof, all the way around.
Lastly, the hardware itself has some question marks. Most modern analog sticks end up suffering from some degree of drift eventually – it has been the bane of the Switch Joycon design for years, but even Sony’s Dualshock and Dualsense designs get it, and Valve is promising they’re trying to prevent it…but that isn’t a guarantee. The dual-touchpads look somewhat unintuitive to me watching the videos – neither matches the screen aspect ratio and it feels like a very weird thing to put in instead of just making a touchscreen, although a touchscreen comes with ergonomic issues of its own in a design like this. The button layout looks a little too flat, the system a little lacking in contouring or shaping to assist in grip for long sessions, and the bottom-mounted grip buttons are cool but also just trying to get an idea of how they’ll feel gives me some mixed emotions. Even on an Xbox controller, with the grip I find best suits me, the idea of bottom-grip buttons feels very odd and I’m not sure how comfortable they’ll be, especially since the shoulder button placement is up high and the whole control array is actually pretty far up as well, which seems like a potential ergonomic nightmare. I just tried grabbing air in a grip that looks like it would work, and it felt pretty bad, not just because I got a weird glance from my wife!
Overall – A Fascinating Piece of Hardware and Reaction
In the end, I’m actually kind of excited for the Steam Deck. Not in terms of actually wanting one, because I know that mobile game systems never really scratch the itch for me – I use my Switch docked about 95% of the time, I put in an embarrassingly small amount of playtime on all of my DS and PSP/Vita models, and the last time I seriously engaged with mobile gaming was as a child with a Game Boy Color and later Advance, but I think it’s a strong contender in the market and it adds a legitimate, well-rounded pocket PC option for people to pick up with a gameplay focus. SteamOS and Proton are big question marks, as is the actual performance of the thing in a wide variety of titles, but it looks promising and I think the demand reflects a lot of real, authentic excitement around it. So many people are fond of mobile gaming and the expansion of PC gaming into that space is really cool.
Will it deliver? Eh, I don’t know for sure. Valve’s track record on hardware is a rightful cause for concern, but at the same time, the most recent launch from them was the Index, and it was very well done. The partnerships they have on this reflect a lot of interest from Valve and a lot of effort put into it, and the choices made are mostly very good. The physical design stuff I question is all hypothetical, because I obviously don’t have a unit in my hands, so it could be very good and comfortable for long play sessions.
For me, it isn’t something I’d buy because mobile gaming is just not where it’s at for me, but I’m not the majority of the market and obviously, Valve has found a goldmine here, provided it’s more Index and less Steam Controller.
If they can deliver on that, well, remains to be seen.
Personally I am on the skeptical side about the SteamDeck and at this point can’t justify buying one (even if it was available in my region of the world)
I share a lot of your concerns, specially about Steam’s track record with physical devices and the performance because of games due to Proton. Even if Valve does have a lot of improvements to Proton that they haven’t shared yet, like they claim, I still don’t see it running games as smoothly as if they were native ports. And this might end up giving a lot of bad PR as someone tries to play a AAA game on it and doesn’t get the performance they expected.
Despite all that, I do hope it succeed. Linux is my system of choice and if it does succeed it has the potential of being a big boom for gaming on Linux. If it doesn’t, which I believe is the most realistic outcome…. then I expect things will stay mostly the same, with only indies and a handful of medium developers making ports to the OS.
LikeLiked by 1 person
I think the popularity of the hardware is a bit besides the point. I use my Steam Controller, Steam Link (both hardware and software app), and Index almost every day. They’re well made devices that continued to be well supported even though two of the three are dead projects.
The “shift” in optics about the potential success of the Steam Deck has nothing to do with the device itself. I think the Steam Deck really only looks sketchy if you’re trying to measure it in terms of being a “service” instead of just a piece of hardware. The 1% of gamers who talk about video games on the Internet are wrapped up in the former, everyone else is more concerned with the latter. A device that will play 80% of my Steam library for the next 5-8 years has concrete value whether the ecosystem around it succeeds or not.
LikeLike