Something No One Saw... But Me
Do you remember Halo 3?
Sure you do. Think back to how that game looked and felt in your mind.
Now, go look at some screenshots or videos of it. Is it how you remember it from your teen years (I'm aging myself here)? Maybe not. The textures are pretty muddy, the facial animations are pretty rudimentary by today's standards, and the environment geometries are pretty low poly. All you Halo 3 enjoyers out there can put your pitchforks away though, because let me be clear: I don't think Halo 3 looks bad per se; its art direction does a ton of carrying here, but the technical stuff doesn't really stand up to modern scrutiny in motion. Graphics aren't everything in video games, and you can't argue with the success of Halo 3, but hear me out.
Now go look at Halo 4. Somehow, this 2012 game was developed on the same 2005 Xbox 360 hardware as 2007's Halo 3. But if you didn't know that, you could easily be convinced that Halo 4 was a next gen Xbox One title, especially when compared to it's older brother. Geometry is denser, volumetric lighting gives everything a sense of depth and scale, textures are crisper, and characters are high-poly and animated much more naturally. I think Halo 4's visuals hold up nicely today, even versus much newer entries in the series.
Both these games targeted 30 FPS - a target which was much more broadly accepted at the time - but Halo 4 managed to achieve largely similar performance to Halo 3, with a higher level of visual fidelity and at a higher resolution to boot.
So what gives? Was there an Xbox 360 Pro? Well, sort of, actually; it was called the Elite, but it wasn't actually more powerful system. It simply had more storage (a whopping 120GB!), and was less prone to suffering the Red Ring of Death. So if there wasn't a boost in horsepower, how could 343 Studios possibly achieve such a drastic uplift in visual fidelity over Bungie's work?
The answer is time, hard work, and a healthy dose of both technical and artistic cleverness.
Send Me Out With A Bang
343 Industries had 5 years between Halo 3 and Halo 4 to gain complete mastery of the Xbox 360 hardware. They were able to optimize the software so completely that it arguably used every possible ounce of power the aging Xbox 360 could muster. New, less resource intensive dynamic lighting techniques and global illumination systems built for their engine did a lot of heavy lifting, and the game enjoyed much higher density geometry due to improvements in the Halo engine's streaming tech.
What the hardware couldn't handle, the developers made up for with incredibly deliberate artistic intent. They played to the strengths of the hardware, instead of trying to force a square peg in a round hole and forcing the hardware to use technologies and systems it wasn't equipped to handle. Much in the way Nintendo plays to the strengths of its hardware when developing its first party titles, so too did 343 Industries. The end result was something of an incredible swan song for the Xbox 360, and a promising new start for the Halo series (we all know how that ended up, but that's a topic for another day).
There's a lesson to be learned from the success of Halo 4, and maybe it seems obvious on its face. However, from my position as a gamer and industry writer, it seems to have been largely forgotten. I think it's worth exploring now through the lens of the current state of the industry.
They Must Love The Smell Of Green
Let's fast forward to today. Literally today: this 22nd day of June in the year 2026 as I'm writing this. The price of the Steam Machine, Valve's answer to the video game home console market, has been announced. Gamers the world over are bristling at the prices, ranging from $1,509 CAD for the base model (512GB of storage, without a controller!?) to an eye popping $2,038 CAD (2TB of storage, with a controller). The embargo on press reviews has lifted, and the prognosis is grim.
Gamers Nexus' Steam Machine Review reports that the system's graphical rendering power is analogous to a Nvidia RTX 3060 12GB, which is a low to mid-range GPU from 2021. If that is to be believed, then this is not a device that's going to go toe to toe with a PS5 Pro when challenged with demanding software like, oh, I don't know, just off the top of my head, Grand Theft Auto VI. The comments sections are filled with despair for Valve and for their wallets. The NoGameOvers Discord server has been discussing how Valve could possibly rationalize pricing an uncompetitive piece of hardware this high.
If you'll permit me to play devil's advocate briefly, it's not entirely Valve's fault. We're all aware of the state of the electronics industry right now. The price of volatile and non-volatile memory are both inflated to a comical degree, thanks to AI datacenters sucking up all the supply to do whatever disruptive and destructive thing they're trying to do. System integrators and gaming hardware designers have an unenviable mandate not only from their shareholders or C-suite executives, but also from their audience: Provide exciting generational improvements on their products while keeping the price down for the average consumer. Unfortunately, it's largely out of their hands. Prices are up, and there aren't any obvious signs that they're going to come down anytime soon. There's not much that Valve or any other video game hardware maker can do to mitigate that pain for their customers.
Or is there?
The history of the video game console market is defined by loss leaders. Here's a non-exhaustive list of video game consoles that were sold at a loss by the hardware manufacturer in recent memory:
Xbox
Xbox 360
Xbox One
Xbox Series S/X
Sega Saturn
Sega Dreamcast
PlayStation 2
PlayStation 3
PlayStation 5 (at launch, anyway)
Nintendo Wii U
Nintendo 3DS (after a price drop)
All of these systems made their money (or attempted to, in Sega's case), by subsidizing the cost of their hardware with software sales. Make good software, and you make your money back on the hardware, and the customers feel like they've received good value.
Unfortunately, the industry seems to be abandoning this practice. These companies want to turn a profit on hardware now. Even Valve, who is uniquely positioned to use the evergreen sales numbers on Steam to subsidize the cost of its shiny new hardware, has elected not to do so, and their language implies that they're trying to have it both ways. Don't take my word for it, though.
Don't Make A Girl A Promise... If You Know You Can't Keep It
From the Steam Machine FAQ:
We think of Steam Machine as an extension of PC gaming, not as a console.
The traditional console model is to sell hardware at a loss and make up the revenue with subscription services or by selling games that are locked-in to the hardware. We think this can make sense for a single business in the short term but that open ecosystems are better for customers over the long term. PC gaming's history proves this: The openness of the PC gaming space has enabled it to be the primary driver of hardware and software innovation for decades.
So then, it's all in the semantics; Valve has made a console-like system, has marketed it like one, and is now rationalizing it being expensive by insisting that it actually isn't one. I agree with them on certain points of this excerpt in a vacuum. Open ecosystems are better for consumers, but that ought not to come at the cost of, well, higher costs. The argument that they would have to have Steam Exclusives doesn't hold water because Steam is device agnostic. Steam runs everywhere, including devices they didn't have to take a loss on, and it generates revenue on those devices too. And you're telling me you can't subsidize your hardware using that incredible revenue stream?
Style yourself as the everyman's gaming saviour, or don't, Valve, but call a spade a spade.
Were It So Easy
It's a double whammy for us as consumers, in the end. Not only have prices of pre-fab PCs and computing devices skyrocketed, but so too have the prices of components themselves, which have traditionally allowed us to actually save some money by cutting out the middle man and building these machines on our own. As it currently stands, a significant chunk of consumers - of the captive audience that these companies were built on - are priced out of any current or future upgrades.
It's not limited just to the Steam Machine, either, but rather to gaming as a hobby at large. Everyone I know went out and bought a Switch 1 during the preorder stage, but I'm the only person in my circle who owns a Switch 2 because most of us feel like it's just not a good enough value at $629.99 CAD - and the price is about to rise by 50 bucks! For my part, I can't upgrade my PC to escape my aging 9th-Gen Intel i9 CPU because parts are just too expensive right now. Looking further afield, the PS5 Pro was already laughably expensive for a home console even when positioned as a premium SKU, and now the less powerful Steam Machine is priced even higher than that.
It's not looking good.
Game development is hard. I'm not going to pretend it's not hard, and I'm not going to pretend that I understand all of it's ins and outs. Developers need, and deserve, tools that simplify and streamline that process for them. Engines like Unity and Unreal are genuine game-changers, pun intended. However, they don't come without their caveats. Unity can struggle with high memory usage, sluggish behaviour on large projects, and problems with system clutter after multiple iterations on their core systems. Unreal Engine 5 can produce incredible visuals but suffers, on current hardware, at least, from performance hitches, shader compilation issues, and a steep learning curve for developers.
The popular solution to these performance problems seems to be to just throw more horsepower at them. Bad frametimes? Get a better rig. Low framerate? You just don't have enough RT or Tensor cores to handle our lighting engine. The fault is (incorrectly) laid at the foot of a discerning consumer who (correctly) considers it unreasonable to need to upgrade their devices annually.
The code that runs the games has to run efficiently on the hardware we've got now. Oblivion Remastered and other Unreal-based games like Hogwarts Legacy suffer from severe frame pacing and unoptimized lighting, and since these problems are at the engine level, they have never been properly fixed. Borderlands 4 launched with performance issues that many felt rendered the game unplayable, and even though fixes have been issued, the damage to the game's reputation is done. The list goes on and on. Remember Assassin's Creed Unity's sorry state at launch? How I muddled through that game at the time, I'm sure I don't know. But you can't just throw more powerful and more expensive hardware at these problems forever.
Companies naturally want to sell you their next big SKU, that's to be expected. They want to sell you their latest and greatest. Indeed, once upon a time, the symbiotic relationship between consumers and tech makers was such that audiences would get pretty dang excited about every new iteration or evolution of a product. The performance uplifts or innovations on display often felt generational, and you knew the prices were going to reflect that, but it would never be too steep a price increase over the last gen offering, and usually the prices would depreciate over time. That is no longer the case however, so a new paradigm will have to emerge between the consumers and the gaming and PC industries.
This brings us full circle, or full ring-world if you like, back to Halo 4.
Wake Up, Chief... I Need You
343 Industries understood the importance of optimizing their engine. It was a labour of love. You can't just throw spaghetti code at a hardware chip and expect it to run efficiently, no matter how powerful that chip is. They slaved over their engine in order to make Halo 4 run as efficiently as they humanly could, and the end result speaks for itself.
I posit that this is the way forward for game developers.
There's almost certainly a significant amount of performance left on the table by developers who don't optimize their products. I don't believe for a moment that we've seen the upper limits of what a system like the PS5 or the Series X can achieve, given what we've seen from studios like Insomniac. Games like Star Fox and the Ocarina of Time remake are coming out on the Switch 2 that are promising to be effective showcases of what a well optimized engine on that hardware can do. Way earlier in this piece I mentioned Grand Theft Auto VI, and that was done with intent; that game is shaping up to be a true litmus test to see what current gen hardware can really do.
To reiterate from earlier, I don't know the nitty-gritty details of game development. But I do know that large shareholder-beholden studios are the first to sacrifice quality for revenue. That's a mindset is more harmful than ever in the current industry environment.
For every well optimized experience, it feels like there's two or three stuttering or crashing disasters coming out of AAA studios these days. The studios responsible (they know who they are) need only look to 343 Industries and Halo 4 for the answer. The current console and GPU generation is getting on in years at this point, so now's the time to take a step back, re-evaluate the backend, and make the Halo 4 analogs of this generation. The Wii had Xenoblade Chronicles. The PlayStation 3 had The Last of Us. The PlayStation 4 had Ghost of Tsushima. The Switch had Metroid Prime 4: Beyond (love it or hate it, Retro Studios stayed on brand and put out a technical marvel). The current generation of hardware could sing, if only the composers would write their compositions with the singer in mind.
Finish The Fight
Games like these that push the system further beyond don't have to be swan songs; the industry, being in an unprecedented hardware cost crisis, is uniquely positioned to make this the new normal. Let's take a break from hardware revisions. Nobody can afford them anyway. Instead, let's slow down, regroup, and master the hardware we already have.
Will we see a Halo 3 to Halo 4 level of uplift? Maybe not. But what we have now absolutely has more to give, and we shouldn't leave it on the table just because history says we have to continue the cycle of buying new consoles and GPUs. To be clear, I'm not saying never; someday it will be time for a PlayStation 6 or an Xbox Series Y or whatever. But the time is not now. For now, let's work on falling back in love with the one we're with. Let's get back to taking pride in the product. Let's slow development cycles down. Let's do more to combat scope creep. Let's hold ourselves to a higher standard, and vote with our wallets, so hardware makers and software makers alike understand that effort in equals results out. That might mean that tentpole games don't come out quite as often, but I believe that this just makes them all the more wondrous to behold when they're the best they can be.
I wonder if Halo: Campaign Evolved will be the Halo 4 of this generation... but running on Unreal Engine 5? I won't hold my breath. The game looks beautiful on the surface, but 343 Industries has gone by the wayside and given way to Halo Studios, and as the third steward of the Halo franchise, it remains to be seen if they're going to be equipped to put the same level of care into the game's backend. There are other candidates though. Time will tell with GTA VI, for instance; we've yet to see that game in motion beyond carefully curated trailers.
But make no mistake: somewhere out there, a game is being made that will show the entire industry what can truly be achieved. A game that resonates with the passion and care that went in to not just the creative side, but the technical side as well. That game is coming.
I don't know about you, but I can't wait to play it.