I’m an Indie Developer, and I Pre-Ordered a Steam Deck 64GB

For years, I’ve wanted to see a mainstream(-ish) handheld PC for gaming on the go. GPD Win and similar Chinese competitors have been paving the way, but their latest models have being putting top-of-the-line ultrabook specs inside, with a high price to match; a great technical feat, but off-putting for non-enthusiasts.

Well, after some sudden rumors of a “SteamPal” device a couple months ago, Valve has announced the “Steam Deck,” a mainstream handheld PC, with a starting price of only $399 USD. Like other hyped-up tech devices, the pre-order rules (starting July 16) were complicated, and it’s possible that the first shipment has already sold out. I was lucky and DID get a preorder in, after a few inconsistent error messages, so I’d recommend you keep trying before the gates open for non-Steam users, or else wait until 2022 when supply and demand naturally balance out, or for the eventual improved v2 model.

(And please don’t buy this on Ebay for a silly price, that makes things worse for everyone and it’s not worth it: you could buy a regular laptop with similar specs in the meantime.)

Anyway, there are different tiers for the “Steam Deck,” the cheapest one coming with only 64GB of old eMMC storage. Virtually everyone is saying to not buy that one (and instead to buy the next tier up for $529). But as a small indie game developer, the 64GB is perfect for me. This post will go over why I’m excited for the device, and why 64GB is OK for my intended purpose.

Steam is bringing portable PC gaming to the masses.

Firstly, the sentiment towards a portable gaming PC doesn’t seem to excite any developer friends I know, or hardcore gamers at large. I don’t know why… maybe they’re used to 4K gaming on a $2000+ desktop, and would balk at anything that couldn’t compete with those specs? Are these the same people who would always choose a PS5 over a Nintendo Switch… and feel a similar gloating when choosing a PC over a PS5? On the other hand, GPD has been succeeding with crowdfunded portable PCs for years, so there’s definitely an audience somewhere (the success of the Switch, compared to its more powerful console cousins, is also proof of this).

I still think there’s a lot of potential for a open-source PC console. The OUYA famously WAS this, but failed, I think due to low-specs and an ARM-based CPU that effectively made it a mobile-only-games-console-but-not-portable machine. Today, even low-powered, but modern, specs can play virtually any game, even the latest AAA ones, if you play at 720p with low graphical settings. My dream machine would be portable, pocketable, and play x86-based games at at least 720p, with USB and HDMI ports built-in. Ideally with a full-size SD card slot for easy mimicking of a “cartridge” system for game deployment.

The “Steam Deck” appears to be this very machine… well, with a few missing things. The system’s odd button placement and 7″ screen makes the device WAY too wide to truly be “portable,” let alone pocketable (I hope future devices from Valve or other competitors… the OS is open-source and Valve is encouraging industry competition here… look at the GBA SP or the PSP Go as more sensible alternative designs, with a screen closer to 6″ for pocket-ability). The “Steam Deck” does have a micro-sd card slot for easy storage expansion, but not a full-size SD slot (realistically, I’d be surprised to see any new laptop to have one either). Early reports aren’t clear, but it seems that the “Deck” will only have 1 USB-C port, a huge disappointment for anyone wanting to treat this as a true “PC” as Valve claims it to be. There will be a separate “Dock” that converts the USB-C to a few USB-A ports, HDMI, DisplayPort and Ethernet, but for a “portable” PC this size, I don’t know why they couldn’t fit ALL of those things directly on the “Deck” itself. It definitely limits people who’re running games off an external hard drive (**I quietly raise my hand**).

The main specs are impressive for the price: a Zen-2 based x86 AMD APU, and 16GB of DDR5 RAM. The APU (CPU + integrated GPU) will be the main bottleneck in terms of frames-per-second, and the RAM will be the bottleneck to ensure a game doesn’t crash when limits are met. Both of these are strong enough to be relatively future-proof for all current, and near-future, PC games on Steam. That’s amazing for the price point of $399: I’d expect a similar laptop to be at least $600. At this price point, I was expecting 8GB of RAM at most, which frankly would have sufficed for minimum settings in most current games, and certainly all older titles.

That price only comes with 64GB of hard-drive space, using a eMMC drive. That’s… way too small. Some of the newest AAA titles take up over 100GB each, so you wouldn’t fit even one of those. There’s some argument that you could still store a couple dozen smaller indie games (or older AAA games), but considering the OS and any metadata already using storage on the device… I think the intent is to store ALL of your games on a micro-sd card if you get the base model. Even at the base price, I would have suggested Valve set 8GB of RAM and a 128GB NVMe SSD for the same price, even if the choice they made will technically allow more newer games to be… “playable”… for a more long-term.

For a marked up price, you can get a 256GB or 512GB NVMe SSD version instead. Those are better, sure. But I’m curious for the games to come, now that the PS5 and newest Xbox use NVMe storage as a standard, promising future games that can load segments of massive open-worlds as you play without pausing (“Beyond Good and Evil 2” might be the only example I can think of). For games that will truly NEED a fast SSD, undoubtedly also requiring a LOT of storage, you might only fit 1 or 2 of them on the 256GB size, and be forced to swap out games frequently. I think the intent for ALL of these tiers is to run your library from the micro-sd card, save for the few games you play frequently, or those few titles that genuinely require the faster storage. And let’s be honest: faster storage is nice, but not “genuinely required” to play virtually anything available today.

(Most of the games I play range between 2 – 20 GB each, 5 – 10GB on average. So the 256 GB model might hold 25 – 50 of my games locally? That’s good for a portable library, but I’d still be swapping out stuff too much for my liking.)

And with a 720p (technically 800p) 7 inch screen, I expect the “Deck” to have no trouble playing even current AAA games at that resolution, with at least 30 FPS. Some of those 100GB+ games we’ll get in the next 5 years? Even if you could technically run them, the small hard drive space kind-of makes those out of the question. Gamers who know those titles would probably laugh at the low resolution and integrated graphics: truly, that audience isn’t the one for the “Steam Deck,” and nothing will convince them to let go of their massive, expensive desktop monsters.

So I bought (tentatively, if nothing goes wrong) the $399 64GB model… why?

On my main desktop PC, I’m already playing most of my games off an external portable 5400 RPM mechanical hard drive. And that’s perfectly fine for me. Sure, initial load times to boot the game are bad for a few large titles, but once I’m playing, loading screens aren’t too bad. I have experienced an NVMe SSD, and it’s great, but it’s not necessary. I’m not a spoiled 14-year-old that thinks 5-second loading screens are unbearable. The 64GB eMMC drive is somewhere in between for speed, and I’m satisfied with that. Of course, I won’t be storing many Steam games on there anyway: as I’ve convinced myself above, I’d just as well store my library on a micro-sd card (a 256GB card can be found for $40, compared to the +$125 upgraded tier), even if loading times would still suffer. Loading times just aren’t a deal-breaker for me.

But playing my Steam game library isn’t why I want a “Steam Deck.” I want it to play MY games. As in, the games I made myself and sell on Steam.

There aren’t that many: I only have 3 titles so far, may or may not have a few more in the years to come. They might take 1 – 4 GB of storage each? 64 GB is plenty. Faster storage would be nice on my own games, but I’ll make do. For me, loading times on my Unity3D games aren’t much improved on NVMe anyway (I should figure out how to optimize my projects better).

The real benefit of that 64GB base model: it gives developers an easy test-bench, to confirm the specs work for their minimum requirements, that their UI is big enough to read on a 7″ screen, and to confirm it runs well on the Linux-based Steam OS. AND MOST IMPORTANTLY: it’s an easy, organic way to take your games to game conferences and conventions.

Running a booth to advertise your game at a convention? You might need to pack a heavy desktop or laptop to give it the best chance to impress. Or you can bring a “Steam Deck” with a USB-C display cable and probably run it fine at 1080p.

Can’t afford a booth, but are attending a convention for fun, or to talk with prospective publishers? You can bring a laptop, propped on a table or your arm, and spend minutes booting it and launching your game, struggling to connect a game controller. Or you can bring a “Steam Deck,” boot your game in seconds, and pass it off to let someone try it. 

My “Steam Deck” could be a great “Dust Scratch” machine. A “SD-DS” machine. Which also happens to play external games when I’m not doing business; but otherwise can be disconnected to not distract a publisher from playing “Rocket League” on my device instead.

For THIS SPECIFIC purpose (showing off projects at conventions, or on trains and planes on the way there), or perhaps for business use and platform testing, the “Steam Deck” 64GB model totally makes sense. And at an exciting price point of only $399 USD? Of course I want one.

It’s a similar reason and logic as to why I bought a base-model Surface Pro 3 device years ago: at the time, the CPU and RAM limits were fine for what I was using it for, and I wasn’t yet sure about the device quality or its place in my life. Years later when it broke down, I bought an upgraded Pro 5 and was happy. Buying the base-model of the “early-prototype” version of a device is like my version of the “early-adopter tax”: years from now, if there’s a new and better version that has the stuff I was missing (more portable, more ports), then I can evaluate if it’s worth upgrading, or if the old version turned out to be unnecessary for my work and play anyhow (ie: the new product category may or may not make sense for me; I won’t know until I have one, even if it’s the lesser version).

My true ideal portable gaming PC, with USB ports and a smaller physical footprint and an affordable price, is probably at least 3-5 years away. But having an American company take this seriously should open the floodgates to make such a device eventually come to pass, whether it’s from Valve, or from Dell, Acer or Asus. In the meantime, the 1st generation “Steam Deck” works for me and what I intend to use it for. I expect to make good use of it exactly up to the point when the better version is available, whenever that may be.