Phew! This is a huge announcement that just dropped. Valve is releasing three new piece of hardware in 2026, and no, it’s not about the Steam Deck 2, at least for now. They are going on three fronts:

• VR with the Steam Frame, a standalone headset that can also work with streaming
• Traditional gaming with the Steam Machine console
• Peripheral with the new Steam Controller (to go with the Steam Machine, Steam Frame, or any computer)

Everything runs on SteamOS, by the way.

Let’s go one by one.

New Hardware

Steam Frame

The Steam Frame is a standalone VR device that runs on an ARM Snapdragon chip, with SteamOS, and runs Windows games. While they briefly mention it, this is biggest news: they have now enough confidence in FEX (the x86 to ARM code translation layer) and Proton to make things work in an actual, commercial hardware.

We have no idea about the performance profile at this stage, and what games will be compatible - but it’s made to run not just VR titles, but non-VR titles as well (showing a huge screen instead of a VR environment). The resolution on your eyes is similar to what you get on a Quest 3, but since it’s a regular PC running SteamOS this opens up a lot of possibilities with what you can do and run on it (especially with FEX).

• Foveated streaming / foveated rendering
• Controllers: no strap, no trackpad, with capacitive finger tracking.
• LCD display, not OLED, as it lets you have faster refresh rates
• No need / no support for base stations, all motion tracking is done on device

The foveated streaming is a key innovation that should enable high speed streaming of frames from a powerful PC to the Steam Frame via the WiFi connection. People who have tested it claim that there was no time where they felt that the flux was getting slower or worse, even when they moved away from the streaming device.

The motion tracking is apparently very, very good as well, and does not require any base station anymore.

Here is a comparison table between the different existing platforms in that space:

Source: vr-compare.com

Steam Machine

This is a small factor box-shaped device, that reminds me of what the Framework Desktop which has a similar footprint. It’s running a semi-custom AMD chip. It features apparently a 6-core Zen 4 CPU clocked up to 4.8 GHz, paired with a semi-custom RDNA 3 GPU with 28 compute units and 8 GB of GDDR6 VRAM.

No info about pricing, but it looks like Valve wants to have a pricepoint to reach as many people as possible. This is what Yazan Aldehayyat said to IGN:

Yazan Aldehayyat: We arrived at this in a variety of ways. We just triangulated our way via different directions. The mandate that we had was, you need to be able to play every game on Steam or at least have enough performance to play every game on Steam at 4K 60 hertz when using upscaling. And so that was the important thing, because we didn’t want people to worry about whether their Steam Machine supports whatever game they’re playing. We just wanted it to be a pretty simple message that yes, if it’s on Steam, this device will have enough performance to play at these settings. But we obviously also wanted to make the device affordable. We understand that affordability is really important, so we kept that in mind and made sure that’s a device that’s going to be reachable for a lot of people.

As to why they decided to develop this machine, there are probably several reasons, but Pierre-Loup Griffais said on the same IGN interview that they noticed that “20 to 30% of Steam Decks are connected at all times to an external display” and considered that there is a market for a console that would be attached to a TV or monitor. There was also quite a lot of community feedback about the desire of such a form factor, so here we are.

Steam Controller

Steam Controller looks like a better Dual Sense. With gyro, nice placed magnetic thumbsticks (TMR) and trackpads on a sweet reachable spot. It has the same inputs as your Steam Deck (with 4 grip buttons at the back), with configurable grip sensors.

It features 4 haptic motors, 2 for trackpad feedback and 2 rumble for game feedback like rumble.

It supports tethered play with USB-C, it also supports Bluetooth connection or the Steam Controller Puck. The Puck is a 2.4GHz wireless connection with polling rate of 4ms. It is interesting that they describe it in milliseconds, as most hardware companies report in Hz. That is about 250Hz. Most gaming hardware would be set at 500Hz or 1000Hz polling rate. In an interview one of the Valve engineers was explaining that since they have control over the whole vertical architecture, from hardware to software, they managed to keep the latency very low, and this even with up to 4 controllers, while a regular Bluetooth connection would get worse as you add a second or third peripheral.

The Puck supports 4 Steam Controllers. I wonder if the Steam Machine would support 8 Steam Controllers with 2 Pucks without lag, one of the biggest problems when I have local party games with my friends with wireless controllers.

The battery life is rated at 35+ hours (8.39 Wh Li-ion), but it should be reduced when paired with the Steam Frame.

Big questions Remain

No Pricing

When the Steam Deck was announced back in July 2021, they had a pricing already cleary on the table. This time, nothing, just vague comments and no clear indication or even range about what we should expect. This does not fare well. If pricing was going to be competitive, they would have already announced it from the get go as part of the big Marketing push. I suspect (we will see if I am proven wrong or not) that this means price will be on the high side. Maybe more than 1000 USD for the Steam Frame, and not far from that either for the Steam Machine. If that’s where they land, they will be fighting for a breadcrumbs of a market. The Steam Deck sold well because it was priced at a very good, considerate range. A home console has a lot more competition - you have the Xbox, the PS5, the Switch 2 when docked, and the DIY PC market too. You can play with the numbers, explaining that you can run 3 bazillions games from Steam, but the purchase price is a real barrier. Go too high and your market will shrink very, very quickly.

My guess is that Valve has several options on the table for pricing, and is looking at the reaction from the audience from this initial marketing push to decide which one to put forward. That’s one way of doing things. Another real possibility is the volatility of prices right now for RAM and NVME drives that would make setting a price several months ahead of release very risky and unpredictable. But then, why do the announce so early?

Why Now?

November is an awkard timing. You may generate some excitment, but there is no hardware shipping in 2025, so you miss the holiday season. Also, it’s not clear why they rushed to present all projects at once. Is there some kind of tsunami coming from another company that we are not aware about? Why the push to talk about all of this without pre-orders? Because a tidal wave of interest on Youtube, X and other platforms is the perfect time to capitalize on the interest and get pre-orders. But there is no timeframe, no clear release date.

It’s very strange. Why not wait for early 2026 with a full announcement where a Buy Now button actually works? This may be related to the fact that they are testing the reception now that the news is out there, ahead of deciding about the pricing and potentially other factors.

Steam Frame: Standalone or Streaming?

It was not super clear from the announcement what to make of the main use for the Steam Frame. Is it mostly made to play games with the cool foveated streaming mode that they spend so much time talking about? Or is the standalone mode really a thing, despite the limitation of the hardware? If streaming is the main use, this means you’ll still need a super beefy PC to run games for the Frame. If it’s mostly made for standalone, there are not going to be a lot of exciting VR games to play with, and then the other use is to play in pancake mode for a huge virtual screen for your existing game library. For the second one, is there really a large customer base for that? The question will become more and more relevant if the pricepoint is high.

Why come back to VR?

As much as the Steam Frame hardware looks cool and exciting, I can’t help but think that… the VR market is dead (I know our own Patola does not agree with me on this!). I see it every time I go to the Tokyo Games Show. Some new headsets, very few new games, and most of them are still rehashes of the same concept or mini-games. It’s very much a small, niche market. And many people who bought a headset tend to reduce their usage fairly quickly after the initial impression. Fundamentally, the Steam Frame changes nothing. It’s still the same VR experience that you get with other options, it will not enable some different kind of immersion. So why are they coming back on the VR market, knowing full well that there’s hardly any audience? Is this a moonshot bet that will take 20 years to become a thing? There are many people like me who choose not to buy VR equipment while being full aware of what it does - and it’s not only about the cost - it’s about not seeing the value.

So you’d think Valve, a software company, would spice things up by announcing a bit VR title at the same time? To make you believe that there’s a point in buying such a system? Well, nothing, at least for now.

It’s great that this looks to be one of the most open hardware options for VR out there, but… that’s not a selling point in itself.

Why the Steam Machine?

I remain a little unconvinced by that choice to launch this piece of hardware. I am pretty sure that 80 to 90% of Steam Deck users already have a gaming PC. And probably, a powerful one for most of them. That’s the whole idea of the Steam Deck, it’s to be able to play on the move (or a separate room) the games you already own on your PC, right? So… who is this new PC for? The current Steam Deck users who have a PC will probably not purchase this. I am also assuming this is not going to be cheap - if we are talking a 900, 1000 USD piece of hardware, that’s not the kind of things you buy to put next to your TV. You can either build a PC for much cheaper than that, or buy a console for that matter. There is probably a non-zero market, but is that really huge? I wonder.

Also, I suspect this will be a non-upgreable piece of hardware. With things soldered on the motherboard. So you may be able to run stuff right now well, but 2, 3 years down the road, it will stop being as adequate and this will become a brick that you need to resell for a better version. They call it a PC, but a PC is fundamentally a sum of its parts: boards that connect to each others. I don’t expect my portable PC like the Steam Deck to be easy to upgrade, because that’s a compromise related to its size. However, a full-fledged PC, yes, I do expect to be able to upgrade things, and not just the NVME drive if you know what I mean. That’s why I game on PC and not on consoles.

So this may be a push to attract console gamers? I don’t see it, though. Steam is powerful in the PC space, but has nowhere near the brand recognition among the other part of the population who defaults to consoles. And this is something I realize everytime I talk to hardcore gamers who are on the console side. They almost have no clue what Steam is, and even if they do, barely show any interest in it. Entering that market means massive marketing campaigns that only companies that sell dozens of millions of units can afford.

We’ll see. I imagine it’s a really good piece of hardware, but at the same time, I’m really not clear on who is going to buy that.

Steam Controller: an Easy Sell

I expect the Steam Controller to sell well (unless they go crazy and price it at 200 USD…) because it’s answering a clear need. Playing FPS games or other mouse driven games with a joystick is not great if you are sitting on a sofa, and having the same controls as what you find on the Steam Deck is the bare minimum these days for a good PC gaming experience on a large screen. It will also nicely replace the aging Steam Controller with several great upgrades.

Repairability

How repairable all of this is going to be? Since that was one of the main selling points of Steam Deck (maybe add this at the end). Should we expect them to sell parts through resellers easily? At least it’s probably going to be difficult to tweak things inside the Steam Frame (apart from the cameras that you can replace in the front).

Going Alone

One of the main outcomes is that Valve has decided to go all in on hardware. There was a time where they tried to corner the market through 3rd parties (in 2014) and they changed gear with the Steam Deck in 2021. More recently it looked like they were trying to convince other manufacturers to jump ship and use SteamOS, with moderate success (the Lenovo Legion Go S). Have they realized that the effort is now futile or too much trouble? With this announcement, it looks like it. A Steam Deck 2 is also clearly on the way sooner or later, so you will end up with a single company that does everything from portable gaming to VR.

Is this a long term bet, or still testing the waters? I can’t fathom the VR headset selling that well, the only question mark is whether there is some appetite for a Steam home (PC) console.

The market will let us know very soon.