The New Steam Machine

The new Steam Machine is expected to land by March 2026. A small, sleek, black cube that Valve say is six times more powerful than the Steam Deck. Is this destined to become a repeat of 2015, or has Valve got enough steam to actually pull it off?

The news of a new Steam Machine comes almost exactly ten years to the day that Valve released the original model in collaboration with various PC hardware vendors. It was essentially a small form factor gaming PC with the very earliest iteration of Steam OS. While - on paper - it was a win, it was far from the PC gaming console the latest version promises to be. Game support was incredibly spotty, Steam OS was in its infancy, and it was as expensive as any other gaming PC. No real surprise it was an utter failure, but it did at least bring with it the original Steam Controller and Steam Link, both of which did a decent job of moving the needle in the lounge-based PC gaming landscape.

But isn’t the 2026 iteration of the Steam Machine also going to be ‘just a small form factor PC' too? Well no. It's actually built and spec'd way more like a games console. Scarily so. They’re sticking with the same AMD silicon stack as the Steam Deck and their building on ten years of hard won R&D to do it.

The landscape was very different 10 years ago when the original Steam Machine landed. There was no Proton compatibility layer, no decade-long investment in software development, no Steam Deck to learn more. Nothing really, beyond the idea and the desire to make it happen.

The new Steam Machine is quite well connected with decent USB options and a built in power supply unit.

Price matters more than performance

It’s a thought process that seems completely out of place for a PC gaming market where performance has mattered regardless of cost. A massive price markup has long outstripped any small spurt of extra FPS on offer for the latest high-end graphics cards. If there’s one thing the success of the Steam Deck can tell us about consumer preference in 2025 though, it's that price actually matters more than performance. Even at launch the Steam Deck was fairly anemic in terms of processing power, but its entry price was incredibly aggressive and the product 'just works'. It has also done a great PR job on the scalability of a lot of mainstream games: you don’t need immense power to perform at lower resolutions. Just take something like Cyberpunk 2077 for example. Of course, it does depend somewhat on what you deem ‘playable’, but since we’re in the PC domain, nine times out of ten we can tweak and perfect our settings to get the balance to our taste.

If there’s one thing Valve has a lot of from their experience with Steam Deck and decades of running Steam - data and analytics. They haven’t been shy in reporting that the specs of the Steam Machine will be as powerful as 70% of the gaming PCs they see in their Steam user base. It’s a numbers game, right? It’s not a sexy number, granted, but for Valve at least, the numbers seem to add up.

The simple small, sleek box design is a winner for the out of office experience - it's ten times easier to discreetly tuck under a telly than a PlayStation 5.

But at what price?

The only number currently notable by its absence is the price. I’m really confident that the price over performance mindset of the Steam Deck is really going to pay off if correctly applied to the Steam Machine. We already know the specs. It’s going to make a large chunk of the most modern and demanding titles playable on our TVs and in 4k. And let’s not forget that it has embedded hardware to make streaming to the Steam Deck (hopefully) the best it’s ever been. But what’s a fair price? Too high (beyond console market norms) and console users will stay exactly where they are, while Valve would simultaneously lose the more hands-on computing punters to their own mini ITX custom PC builds. If they can get the price just right, though, then there’s little point in anyone with a Steam account waiting around for Microsoft and Sony to make a move - they’ll be free to jump straight in.

If the bill of materials truly demands a high RRP, Valve would do well to invest in their ‘console market move’ in the most traditional of ways - make a loss on the hardware to keep costs down in the short term and earn it back in software (at least until the bill of materials finally fits their price). Valve can afford it - I guess we’ll see how serious Gabe is about a full-frontal assault on the console market. Personally, I’d pay 500 with a new Steam Controller (heck, I might pay 550). 599 and up is running well into dangerous territory. 699 or more would be a complete bust.

A simple decision

As an existing PC gamer the decision to buy comes down to this: Do I want to spend a decent chunk of change on an out-of-the-box lounge solution, or spend around the same on a new graphics card for my ailing office-based PC? I’m at a stage in my life where the former is looking way more attractive. I’m old, lazy and I have two young children. A sleek cube in the lounge that I can play directly or stream to my Steam Deck? Take the money. I guess Valve is also banking on that all important 70% of their existing market to feel the same way. What about you?