The beginning of a new console generation is always a big event. Brand new powerful consoles, setting a new bar for the kinds of experiences we get to enjoy as gamers. While returns may be diminishing with each generation, there’s always more that can be done under the hood to make experiences better. But as power skyrockets, so too do the prices we as consumers have to pay. On the surface, a console such as the Xbox Series S, is a blessing because it’s more affordable by a large margin compared to its larger sibling the Xbox Series X and more than likely its rival the PlayStation 5. But that lower cost of entry comes at the expense of the consoles power, and by extension, lowers the bar for next generation games.
Back in March, both Microsoft and Sony revealed the technical specifications of both their next generation platforms. At the bare minimum, what developers had to work with was an RDNA2 GPU(Both Consoles), 13.5GB of GDDR6 VRAM(Xbox Series X), a GPU clock speed of 1.825GHz(Xbox Series X), 10.28TFLOPs(PlayStation 5), Memory Bandwidth of at least 336GB/s(Xbox Series X), 36 Compute Units(PlayStation 5), CPU clock speeds of at least 3.5GHz(PlayStation 5), 8 Core/16 Thread Zen 2 CPU(Both Consoles), and SSD I/O speeds of at least 2.4GB/s(Xbox Series X).
These are absolutely fantastic numbers that made developers jump for joy because it gave them so much more to play with in terms of power. In the current generation of consoles, developers are bottlenecked mainly by the Xbox One/Xbox One S, but also by the HDD storage medium and anemic Jaguar CPU’s found on both the Xbox One and PlayStation 4. Even their mid-generation refreshes couldn’t escape the shackles placed on them by the initial launch platforms. And now developers find themselves facing a similar conundrum with the Xbox Series S. So lets see how the numbers stack up.
13.5GB of GDDR6 -> Reduced to 8GB of GDDR6(Realistically, potentially 7.5GB.)
336GB/s of Memory Bandwidth -> Reduced to 224GB/s
GPU Clock Speed of 1.825GHz -> Reduced to 1.565Ghz
36 Compute Units -> Reduced to 20 Compute Units
10.28 TFLOPs -> Reduced to 4 TFLOPs
CPU Clock Speed of 3.5GHz -> Reduced to 3.4GHz
SSD I/O Speeds of 2.4GB/s -> No Change
8 Core/16 Thread Zen 2 CPU -> No Change
RDNA2 GPU -> No Change
When third party developers begin developing their titles, they select their platforms. The weakest platform on the list becomes their minimum specification. Anything they do with the game must be capable on this platform. They then scale up accordingly for the more powerful platforms, making sure they don’t add anything to the game that the lowest common denominator cannot handle. The less power they have to work with as a baseline, the less they can do with the game. It’s pretty straightforward. Already, developers have begun to publicly complain about the Xbox Series S, a console that threatens to bottleneck what next generation means. By throwing a bone to those who can’t afford the more premium consoles, Microsoft has effectively given not only their prized Series X a self-handicap, but also by extension the PlayStation 5 at least as far as third parties are concerned. This also lowers the bar for next generation PC gaming. As many games are developer with consoles in mind, the PC versions typically don’t push the hardware as far as it can actually go. You can read some developers thoughts on the matter here:
As you can see, the reaction is actually more mixed than one might expect. There are developers genuinely concerned and others that aren’t bothered by it. Part of it likely comes down to personal experience and/or capability. Some developers are just better than others at optimizing even the hungriest of games while others struggle to get even their simpler games to run on the strongest hardware. And when you’ve been accommodating aging PC hardware for so long, there’s actually not really an excuse for the Xbox Series S. But you also have to ask yourself: What is really the next generation experience? The specifications of the Xbox Series S are honestly extremely telling: Microsoft firmly believes that the GPU is just a side show and the real stars are the CPU and SSD.
Look at the games we’ve gotten this generation, look at how large and dense these game worlds are. Despite the limitations of the Xbox One holding them back, developers achieved massive strides in rendering huge worlds and beautifully detailed characters. In the next generation, the GPU will see the lowest returns, it has piqued. But the CPU? It has now caught up to high end desktop processors and that is consistent across all three next generation platforms. The Xbox Series S is only .1GHz slower than the PS5's CPU and that’s only when multithreading is activated, otherwise it’s actually still higher at 3.6GHz. It’s ability to process things like AI and other instructions is equal, barely any compromise at all was made because Microsoft deemed this critically important. The same can be said for the Solid State Drive, it retains the exact same specifications as the Xbox Series X at 2.4GB/s of Raw throughput and 4.8GB/s compressed(It has also been stated to be higher than this, with the example of 6GB/s given, however Microsoft opted to advertise a conservative estimate so as to not overpromise.). This is still far higher than a 55MB/s HDD and more than enough for an engine such as Unreal Engine 5 and its Nanite technology(And yes, it is fully capable of running on Xbox Series S and Xbox Series X. The only reason it “can’t” is because it was developers specifically for the PlayStation 5. If ported over and optimized for the Velocity Architecture rather than just straight raw hardware, it would function just fine and the same can be said for PC as well which is now a reality thanks to NVIDIA RTX I/O.). The SSD is the key feature of the next generation, allowing instant loading and seamless transitions with no perceptible cuts, no tricks.
So why so much fuss over the GPU? Developers are used to working with limited resources, and having experienced similar situations myself, I can understand how hard it can be initially to realize that you can do so much more with the right tools. For one thing, the Series S will not be running games at 4K, instead it will run them at 1080p or 1440p, this drastically reduces the resolution that needs to be rendered as 4K is expensive. Saving memory by cutting back on that frees up more for other things such as textures and geometry. Then, when you throw in things such as Variable Rate Shading and Sampler Feedback Streaming, you continue to reduce how VRAM you actually have to use at any given moment, and that’s not taking into account compression and the speed at which textures and objects are being loaded in and out which is near instantaneous thanks to the SSD, meaning the VRAM utilization will constantly be shifting. There’s so much under the hood work that’s been done to drastically reduce the resources needed to render games at high fidelity, and yet it feels like some developers haven’t yet realized that they can now do more with less as far-fetched as that sounds.
Now, I don’t intend to sound condescending or like I know more about how this shit works than they do. I don’t make video games(Though I wish I did.) or develop the technology that makes them function. I’m merely looking at the data that has been presented to us, and it just seems odd that despite all of these advancements that completely change the core foundation of how games are made and how they’re visualized on screen, some developers don’t see anything different and feel like they’ll run into the same problems they always have because of the Xbox Series S, even though they seemingly didn’t question the Xbox Series X or PlayStation 5 which feature the exact same technology. I’m not saying it can’t cause any issues at all, but I also don’t think it will dramatically change what is capable in the next generation either because developers have already hit the proverbial ceiling on what a GPU is needed for, and it’s now more about what an unshackled CPU can do and the game changer that is the SSD. Once the current generation consoles are dropped, and developers have had more time to develop games on these platforms, I do feel like tunes will begin to change. But only time will truly tell, because for all I know, problems that Microsoft and Sony think they’ve solved, could appear again as developers push the consoles to their absolute limits, and those developers who are already concerned, will feel vindicated that their fears were not entirely unfounded.