Let us begin, children, with a simple question. What is a console?
Go on. Take a moment. It sounds easy. It’s not easy. It is, in fact, a question that starts reasonable and becomes increasingly destabilising and irritating the longer you sit with it, like trying to define whether a hot dog qualifies as a type of sandwich.
You start with a confident answer. The confident answer develops holes. The holes get bigger. By the end you are staring at your PlayStation 5 with the unsettling feeling that you have been lied to, not maliciously, but consistently and at considerable scale, for approximately thirteen years (and why are my hands covered in ketchup?).
The occasion for this existential spiral is Project Helix, the codename for Microsoft’s next Xbox, confirmed at GDC last month with the kind of careful understatement that large companies use when they are about to say something significant that they’d rather didn’t sound too significant. Project Helix will run on Windows 11. It’s rumoured it will play PC games from Steam, Epic, GOG, and wherever else you have accumulated a digital library of things you own but never play. It will also, of course, play Xbox games.
It will cost, by current estimates, around £850, which is the price of a reasonable entry-level gaming PC and which Microsoft is hoping you will read as a coincidence.
It is not a coincidence. Project Helix is a PC. And the interesting thing is not that Microsoft built a PC and called it a console. The interesting thing is that they are barely even pretending otherwise, which makes them the first console manufacturer in thirteen years to extend this basic courtesy to the customer.
But before we get to Project Helix, we need to answer the question. What is a console? Because the answer depends entirely on which decade you are standing in and which manufacturer you are looking at. And unless your name is Nintendo, there is very little to physically separate your system from the humble PC, other than a great deal of carefully managed spin about what is actually powering your plastic brick of choice.
Definition One: A Console Is Special Hardware That Does Things a PC Cannot
This was the correct definition from approximately 1977 to 2012, and it was genuinely true, and it made consoles interesting in a way that had nothing to do with marketing.
The Atari 2600 had a custom chip called TIA that handled graphics and sound in ways desktop computers of the era simply could not replicate at the same price. The NES had bespoke silicon that extracted 8-bit magic from constraints that would have floored a generic computer. The SNES had the Super FX chip, which developers literally soldered into cartridges to add 3D rendering capability the console was never supposed to have. The PlayStation had the GTE, a geometry transformation engine that made early 3D possible on hardware that cost £299 in 1995. If any of this is making you nostalgic, our retro PC games list will finish the job.
These were not PCs with a different logo on the front. These were architecturally distinct machines that required specialised knowledge to program, that could do things equivalent desktop hardware could not, and that justified their existence as separate categories of device on purely technical grounds.
The PlayStation 3 carried this tradition to its magnificent, maddening extreme. The Cell Broadband Engine was a processor co-developed by Sony, IBM, and Toshiba that bore essentially no architectural resemblance to anything in a desktop PC. It had one primary processing core and seven Synergistic Processing Elements running in parallel, a design philosophy derived from supercomputer research, and a learning curve so severe that developers spent three years writing letters of complaint before producing games that justified the whole miserable exercise. Naughty Dog’s work on Uncharted 2 and The Last of Us on Cell hardware remains one of the most impressive feats of programming ever committed to consumer silicon. The machine was genuinely exotic. Genuinely different. Genuinely a console, in the truest sense of the word.
The PS3. Curvy, exotic, and utterly impenetrable. Seven Synergistic Processing Elements of pure drama. (Image credit: Sony)
Then the PlayStation 4 arrived. AMD x86 processor. AMD Radeon graphics. Standardised memory architecture. The Cell Broadband Engine was retired without a funeral and everyone quietly agreed not to discuss it at family occasions. The PS4 was, beneath its matte black exterior and carefully cultivated air of exclusive mystery, a mid-range gaming PC from 2013. A good one. A well-curated one. One that came with a very nice controller and excellent exclusive games. But a PC. Recognisably, undeniably, structurally a PC.
The Xbox One was also a PC. The Xbox Series X is a PC. The PlayStation 5 is a PC. They are all PCs. Under the bonnet, beneath the boot sequences and proprietary stores and carefully licensed storefronts, there are x86-64 processors doing x86-64 things and AMD graphics cards rendering AMD frames. The jig, as they say, has been up since 2013. Nobody has wanted to say it directly because saying it directly is awkward for companies selling you the dream of a dedicated gaming machine. But there it is. You deserved to know.
So if Definition One no longer applies to Sony or Microsoft, what are we left with? We are left with Nintendo, who have been watching this conversation from across the room with the expression of someone who knows something the rest of us do not.
Definition Two: A Console Is Whatever Nintendo Is Doing, Which Is Completely Different and Always Has Been
Here is where the argument gets interesting. Here is where one of us, in the internal debate that produced this article, had to concede a point, which was deeply unpleasant and will not be happening again.
Nintendo is not making PCs. Nintendo has not been making PCs. Nintendo is currently making something that is architecturally closer to a high-end Android tablet than to either a gaming PC or a PlayStation 5, and this is not an insult. It is a description. The Switch 2 runs on a custom NVIDIA chip built on ARM architecture: the same fundamental instruction set as your smartphone, your tablet, and approximately every piece of mobile computing hardware manufactured in the last decade. It has 12GB of LPDDR5X memory, which is laptop RAM, designed for low power consumption rather than raw bandwidth. The whole thing weighs 415 grams, which is about the weight of a very determined sandwich, and it can be lifted with one hand because Nintendo designed it to be lifted with one hand.
Nintendo is doing something genuinely different. Not different in the way that Sony and Microsoft do different, which is different marketing materials for the same fundamental hardware decisions. Differently different. Architecturally, fundamentally, from-the-ground-up different in its priorities and its engineering philosophy.
Nintendo play by their own weird set of rules, but it works (Image credit: Wikipedia)
The reason is straightforward and Nintendo has been consistent about it since approximately 2006: they are not trying to win a specs race. They gave up on the specs race when they looked at the costs involved in competing with Sony and Microsoft on raw performance terms and decided, with the calm clarity of a company that had been making games since 1889 and understood perfectly well what actually mattered, that they would rather make something people wanted to carry on the bus. The Wii was underpowered and sold 101 million units. The Switch was underpowered by contemporaneous standards and has sold over 150 million units. Nintendo has been wrong about many things over the years, but they have been comprehensively, expensively, embarrassingly right about this one.
The Switch 2 is therefore the only current major gaming platform that can legitimately claim to be something other than a PC in different packaging. It is a mobile device that sometimes connects to your television. It is a handheld that periodically pretends to be a home console. It plays Mario, which is the single most compelling exclusive software argument in the history of gaming and renders approximately every other platform comparison irrelevant.
If you ask whether the Switch 2 is a PC in fancy clothes, the answer is: no, it is a phone in fancy clothes, which is not the same thing, and which somehow makes it more honest than everyone else while simultaneously being very strange. Nintendo has no definition of a console. Nintendo has a definition of a Nintendo, and it works, and we should leave them to it before they do something unexpected with a new peripheral.
Definition Three: A Console Is a Managed Experience Delivered by Hardware That Is Definitely Not a PC, Please Stop Asking
This is where Sony lives. This is the definition they would offer if you produced a microphone in their direction, except they would not use the words “PC hardware” because that would undermine twenty years of brand positioning, so they would say something about “custom silicon” and “the PlayStation experience” and gesture firmly away from the AMD logo embossed on the chip inside the machine.
The PlayStation 5 is a genuinely excellent product by Definition Three. The boot experience is fast. The DualSense controller is the best input device ever attached to a games console, and we will have that argument with anyone. The first-party games are solid to exceptional. The user interface is clean and logical and mercifully free of anything that looks like a Windows dialogue box. When it does not work, which happens considerably less often than any PC, it fails in predictable ways resolved by predictable means. This is the promise of the curated platform. Sony builds the garden. You play in it. The garden is very nice. Several of the best PC games ever made arrived on PlayStation first, which tells you everything about how the garden is tended.
The PS5 cannot play PC games not because the hardware cannot run them, but because Sony has chosen not to allow it
The catch, which Sony would prefer you not to examine before purchasing, is that the garden runs on x86-64 soil with AMD bedrock and RDNA roots. Everything growing in it could, in principle, grow in your PC garden too. The exclusivity is not technological. It is contractual, curatorial, and commercial. Sony has decided what you can play on their hardware, and the decision is enforced by business relationships and store policies rather than by anything the chip itself requires. The PS5 cannot play PC games not because the hardware cannot run them, but because Sony has chosen not to allow it, which is a very different thing and worth being clear about.
This definition of a console is therefore: a PC, managed by a company with strong opinions about what you should be doing with it. That is still a legitimate value proposition. Plenty of people want someone to have strong opinions on their behalf. Interior designers exist for exactly this reason. The managed experience is not worthless. But it is a customer service philosophy delivered by x86 hardware, not a genuinely distinct category of machine. The exotic beast has been replaced by a well-organised filing cabinet. The filing cabinet is excellent. It is still a filing cabinet.
Definition Four: A Console Is Whatever We Say It Is, and We Say It Is a PC Now
Ladies and gentlemen, Project Helix.
Microsoft confirmed at GDC that its next-generation Xbox runs on a custom AMD chip with RDNA 5 graphics, 48GB of GDDR7 memory, and Windows 11 as its operating system. It will play Xbox games natively. It will play PC games from Steam, Epic, and any other storefront that cares to show up. It will have an Xbox Mode interface sitting over the top of Windows and making it look like a console dashboard, in the same way that putting a bow tie on a labrador makes it look slightly more formal while remaining fundamentally a labrador.
Jason Ronald, Microsoft’s VP of Next Generation, was admirably direct at GDC: “The days of people defining themselves as console or PC gamer don’t really exist anymore.” This is technically an observation about player behaviour. It is also Microsoft quietly saying: we are done pretending. We are a software company that makes Windows. We are going to sell you a Windows machine. We are going to put a nice interface on it and call it an Xbox. You are welcome.
Project Helix is, therefore, the fourth definition of a console: a PC that has stopped calling itself a PC because the word “PC” does not shift units in the living room market. Except this time, unusually, Microsoft has stopped hiding it. The curtain has been pulled back. The wizard is visible. The wizard is running Windows Update.
The Awkward Part: Will It Actually Work?
Here is where optimism and track record have a difficult conversation.
The console player seeking a simple, managed, plug-in-and-play experience is getting a Windows machine. Windows is not simple. Windows is a forty-year operating system that contains multitudes, not all of them welcome, including a significant number of background processes, update mechanisms, and decisions that surface at inopportune moments with the specific energy of a cat sitting on a keyboard during an important meeting. Microsoft’s Xbox Mode layer may smooth this into something approaching console serenity. The history of Windows on non-traditional form factors suggests cautious scepticism is the appropriate response, and we are nothing if not appropriately sceptical.
The PC player, meanwhile, looks at a $999 fixed-spec machine and asks what exactly it offers over building their own system at equivalent cost, with the flexibility to upgrade individual components in two years rather than buying a whole new box. The answer may be genuinely compelling for players who want a living room machine without the cable management theology that PC gaming occasionally requires. It may also not be compelling enough when the alternative is an unapologetic PC that does not have a console identity crisis twice a year.
Look! It’s a console! It’s a PC! It’s a…hybrid! Who the hell knows. (Image credit: Microsoft)
The worry, stated without diplomatic softening: Project Helix could end up being the worst of both worlds. Not the best of both worlds, which is the marketing pitch, but the worst, which is the risk. Console players may find Windows lurking under the Xbox Mode veneer at precisely the wrong moment. PC players may conclude that a managed ecosystem at $999 is a significant amount of money for flexibility they are not being given. The middle ground between console simplicity and PC capability is genuinely difficult to occupy, and the history of devices that have tried is not universally encouraging.
Valve is attempting exactly this with the Steam Machine, having already done it partially and successfully with the Steam Deck, and Valve has spent a decade building SteamOS into something that actually behaves itself in a living room context. Microsoft is proposing to achieve the same result using an operating system with a rather more complicated relationship with the concept of staying out of your way. The Zune was, in its time, a perfectly good MP3 player. We are merely noting that history exists.
Additionally, Microsoft’s recent stewardship of the Xbox brand has contained enough misfires that any announcement from them arrives with a reasonable expectation of scepticism attached. The gradual dissolution of console exclusivity. The Game Pass economics. The Activision Blizzard acquisition followed by the layoffs. The general impression of a platform that was not entirely sure what it was for or who it was for. New Xbox CEO Asha Sharma has described Project Helix as part of the “commitment to the return of Xbox,” which is, again, the kind of language you use when the brand went somewhere it needed to return from, and the candour about that is genuinely appreciated. And a reminder that the RAM shortage currently pushing prices through the roof is not going to make a $999 device any easier to swallow.
The Verdict
So. What is a console?
In the original sense: a purpose-built machine with architecturally distinct hardware that can do things a PC cannot. That console is dead. It died in 2013 and Nintendo was the only one at the burial, and Nintendo promptly left to go and do something completely different with mobile chip architecture, which is very Nintendo.
In the modern sense: a managed PC experience designed for the living room, with someone else’s decisions about the software environment built in as a feature rather than a constraint. That console is alive, doing extremely well commercially, and is called a PlayStation 5.
In the Project Helix sense: a PC that has decided to be honest about it, for better or worse, at a price that will make you sit down quietly for a moment.
The most truthful version of a modern games console, it turns out, is the PC you already own, or the one you could build for the same money. But Project Helix is the first console to agree with you on that point out loud, and there is something almost refreshing about it, in the way that it is almost refreshing when someone finally admits they have been doing something you have known about for years.
Whether almost refreshing is worth $999 is a question we will answer definitively when the thing actually exists and we can tell you in a review. Until then: the console is mostly a PC in fancy clothes. It always was, since 2013. Project Helix has simply arrived at the party without the fancy clothes on. If you want a flavour of how GPU manufacturers have been playing fast and loose with the truth on performance for years, this particular honesty from Microsoft is more remarkable in context than it might appear.
As for Nintendo? They’ve been dressed batshit crazy since day one. And that is fine, we all have that one weird friend we like to indulge once in a while.
Project Helix is expected to launch in 2027. Estimated price: $999 / £850.