The Game Industry Update for Jan’25

Bringing you the definitive opinion on the industry since 1996

Hi. Hello. So, normally I try to write with a level of intellectual rigor, but occasionally I like to just express my gut opinions on things. The Gaming Industry is a particularly large target for my opinions as I’ve spent so long analyzing and participating in it, so many of these off the cuff articles will be about it.

I will relegate these kinds of articles—along with my reviews of various media—to the “Crow’s Opinion” section of Ars Corvi for you all to subscribe separately and at your own discretion.

---

The Switch 2 was officially announced and people proved that they don’t deserve it

Finally after 7 years, 10 months, and 13 days of everyone and their dog providing “legit” leaks about Nintendo’s next console generation, culminating in a year of NDA breaking and skirting even the CCP could only dream of orchestrating, the Switch 2 has finally been announced in a brief teaser for a full reveal in April.

The collective consensus: there were no surprises.

Of course there weren’t. The final design proved that the leaks going around weeks ahead of time were accurate. Likely confirming that a lot of other claims made are accurate. I would dare say that Nintendo only went ahead and announced the thing now just because of how accurate they have been.

Take for instance, this article from The Game Post on January 11th on the Genki mockup that was floating around CES. Note the images used by the Game Post here of the “Switch 2” which came from that and compare to what Nintendo finally unveiled:

Image by Genki, sourced from The Game Post

Image by Nintendo, sourced from their site

Yeah…

I still think the analysis on its graphics performance has been speculative at best, but I think it’s safe to say that the rumors leave little else to the imagination. Backwards compatibility was confirmed and they did a little swoopy animation with the controllers over a surface implying that the mouse-like pointer functionality rumored on the updated Joycons is also going to be a thing.

Personally, I’m very happy with this announcement. The Switch has proven to be a versatile and effective gaming platform. Its successor being the same thing but with more compute power and some other small architectural modernizations will be a huge plus for developers and players. All Nintendo’s platform has been lacking is that: power. And frankly, I applaud them for never giving more than necessary in light of my realizations on how often us software people abuse the power we get.

Some people are less enthused, however:

https://x.com/TRIPSTER0/status/1880424506151334260

For the record, Tripster is slightly misrepresenting Max’s opinion here. Max only calls the announcement a Nothing Burger. But the rest of the sentiment does seem to match what Maximillian Dood is saying in the clip.

I’m not throwing shade on anyone involved here, but I do want to point out a particular opinion that I do want to point out as silly. An opinion that a significant amount of investors held to, as the Wall Street Journal reported.

I 100% agree the Switch 2 may have no surprises. What I disagree with is that this is a disappointment.

We didn’t need any innovation, we just need an even more solid base to build from

The days of constantly revolutionizing the gaming world with consoles are long behind us. In fact, I’d personally argue that they ended with the Oculus Rift. Short of wiring people directly to machines (a thing we really may not want to do for the sake of a hobby), you can’t really get more immersive than VR. Especially VR that’s connected to the internet. And that’s the only frontier of gaming left that’s really dependent on new kinds of hardware.

As I’ve talked about before (how many times can I link to myself, I wonder), there’s very little on the core system of modern computers that’s holding us back from creating any sort of gaming experience we imagine. Almost all of the innovation left to do is cultural. We’ve experimented with a LOT of different ways to play games these last 60+ years, and Nintendo has already dabbled in pretty much all of them. A lot of things are possible but not a lot of them are things that people want to do.

That’s not to say they don’t have their audience, but the audience’s can’t sustain the kind of gaming ecosystem at the core of Nintendo’s business. Their consoles have to have broad appeal. The Wii U would have been a success for a smaller company. Nintendo is not a small company.

Thus, Nintendo has to sell a console that’s general purpose. And the first Switch already fit that bill with flying colors. Short of AR/VR, which Nintendo purposefully avoided [outside of their LABO experiment thing] *correction from my Nintendo Expert editor, what is it not at least theoretically capable of? I won’t repeat things I already talked about in my full article on the subject of console releases. Instead, I’ll just reiterate the following.

We don’t need new hardware novelties. We haven’t even finished figuring out how to use the hardware we have!

No one should have expected anything different

It seems like a lot of people missed the metanarrative about Nintendo that was laid out with the launch of the Switch: their Wii is dead. Long live the handheld division.

While everyone had somehow convinced themselves that Nintendo’s console strategy revolved around crazy innovation (a view that was frankly overblown if you remove the Wii and its successor from the equation), the handheld side of their business was holding strong with their real hardware tactic: gradual iteration.

• The GameBoy took the Game & Watch and added the ability to play multiple games on one device and added an external data tether

  • This was pretty revolutionary

• The GameBoy Color… added color

• The GameBoy Pocket… made it smaller

• The GameBoy Advance gave it more processing power and added more buttons… to match the already released Super Nintendo

• The GameBoy Advance SP made the screen light up and the device fold in a clamshell

• The GameBoy Advance Micro… made it smaller (but skipped the folding)

• The Nintendo DS gave it a second screen with a resistive touchscreen, a wi-fi transceiver, more processing power, and dedicated system data storage.

  • Overall, quite a significant jump

• The Nintendo DS lite… made it smaller

• The Nintendo DSi added cameras

  • The digital storefront and expanded operating system is significant, but also that was mostly software changes by that point

• The Nintendo 3DS… made the screen 3D

• The Nintendo 3DS XL… made it bigger

• The Nintendo 2DS… made the screen not 3D

• The New Nintendo 3DS added a little nub[, zR and zL buttons, and more of that sweet, sweet computing JUICE] *correction by my Nintendo Expert editor

• The New Nintendo 3DS XL… made it bigger

• The New Nintendo 2DS XL… made the screen not 3D again

The Switch then came in and merged in the better of the features from the console side of things. Even the docking is really just an iteration of what the Wii U was doing. It seems really dramatic because these two threads merged—in fact, the divisions themselves merged. But, note that by this point in history people were already connecting controllers to their iPads to play games commonly enough that companies were making and selling products specifically for this. So, a super reductionist view could look at the Switch and say that all Nintendo did was provide a more polished version of what was already becoming a gaming cultural norm.

Now, looking at it that way fast forwarding to today, what was Nintendo’s other option for a new console? VR.

Well, Oculus and PlayStation are already taking their spot there on both the portable and high-fidelity ends of the market—and that’s just on the budget side. Additionally, that market has existed since 2016 with some fairly active investment and still does not move the numbers a company like Nintendo needs to justify its place.

I should also add, though I’m unsure how true this still is, Nintendo is a “blue ocean” company. They don’t like moving into fiercely competitive markets. The Switch was a once in a lifetime opportunity in that instance where the concept was not actually that novel but the market was severely underdeveloped even though it did exist. You can thank Apple for that and their consistently terrible understanding of the gaming market. They had the hardware, but they do not have the culture.

Well… I think I’ve ranted on Nintendo enough for now.

tl;dr The Switch 2 looks to be exactly what Nintendo should be releasing and anyone expecting more should get their heads out of the clouds and realize we have a beautiful landscape of possibility as it stands.

Nvidia is shouting “AI” from the rooftops

To NO ONE’s surprise, Nvidia focused all of their 50-series GPU related announcements at CES on AI related features. They will once again be the top dog of power consumption and proprietary tech that uses novel methods to render games. Of particular interest to me was the Neural Material features which essentially extends the kind of image generation functionality of their Machine Learning processes to all layers of a “material” in 3D rendering. That is, the ML algorithm not only determines the final pixels of the underlying low res texture, but also the layers that determine more specific lighting properties like subsurface scattering.

This is actually really really cool. I don’t know how practical it will end up being for actual game development, but like raytracing it can actually allow for certain effects to be rendered which we otherwise do not yet have performant traditional algorithms for.

Dampening the excitement for this is the fact that since it is Nvidia exclusive, game developers will still have to develop and ship traditional materials for everything, so it won’t effectively save any storage space and the actual performance characteristics of the technology are unknown. But the results they’ve shown off are impressive enough that even just a handful of releases showcasing the feature will make the fidelity crowd quite pleased with getting a 50-series card.

Multi-frame Gen I was more excited about until I realized that they aren’t actually doing frame exrapolation just more interpolation. Having recently used AMD’s FSR 3 frame gen and actually having a good experience, I don’t doubt Nvidia owners will appreciate that much more fluidity (especially considering that companies are now producing monitors with refresh rates of 700+, a frankly pointless exercise, I think, but might as well let people try it I guess). However, since it still has the same fundamental latency behavior, this is only useful for people with monitors reaching the 240hz+ range as you need a game logic frame time of 16ms (60fps) or less to consistently mask the downsides of the interpolated frames, which means the game can’t take any longer than that to produce each of the “key” frames and generate the interpolated “tween” frames.

I think using animation terms will be far more fruitful in discussion than saying things like “fake” and “real” frames. They’re all fake.

That said, there are two things I was excited to hear about from that whole endeavor:

1. They developed hardware level frame pacing technology! According to their presentations, multi-frame gen is very pacing sensitive so they finally put their weight into tacking that dragon with the resources only they can as the silicon designers. If this proves successful, then we may be looking at a new definitive solution. We shall see, but just the hope of it is exciting enough for me.

2. They fundamentally updated the algorithm behind DLSS from convoluted neural networks to a “transformer” based algorithm. Frankly, I don’t care enough about the specifics of Machine Learning to even try and explain the difference to any of you, but the point is that we can actually expect a serious update to the image quality and performance of DLSS—which is already the most impressive upscaling technology out there. So, hitting that 16ms frame time will presumably be more achievable with DLSS 4 than 3. At the very least, it should look prettier when you do.

Also of note from Nvidia is that their cards did not significantly increase in price. Now, some people are complaining that the 5090 is more expensive than the 4090 MSRP… but come on. That’s a tier of card that was never hiding the fact that their are luxury products.

If $200 dollars makes the difference between whether you will or will not buy a 5090, you should not be buying a 5090. This is frankly almost true for the rest of the stack, but the 5080 and below are at least in the range where your average American could find a way to fit it into their budget but have to decide how they’ll balance between CPU and GPU.

Meanwhile, AMD is looking to make that entire question of value irrelevant without even saying anything

There’s been much speculation around why AMD didn’t give any performance details whatsoever about their two new “9070” series cards at CES. To me, the answer is quite obvious. Nvidia is so dominant in the high-end space that you don’t really have any good option aside from letting them announce their plan and then tailor your message around them. It’s not like the company as a whole didn’t get any limelight. When it comes the CPU market, AMD is starting to enjoy the same luxury of dominance Nvidia has with GPUs.

And to that end, while the naming schemes in the CPU space are absolutely atrocious, the Strix Halo APUs are really exciting. I’m actually hoping for high-end “System on a Chip” processors to become an option in the desktop space. Theoretically, we could start to see the advantages of gaming console architectures on the PC platform as SoC’s allow for unified memory between the Central and Graphics processing. But, that’s all still just hypothetical.

What’s also hypothetical, but so widely leaked that it’s likely true, is that the 9070 GPUs will have performance near AMD’s current flagship card for just over half the cost. In addition, FSR4 is confirmed to bring ML based upscaling to the platform, and the new card will actually have serious hardware accelerated raytracing, bringing them up to a level that makes Nvidia’s low-tier cards a much, much harder sell.

As much as I promote raytracing and path-tracing as the way forward for rendering, the practical reality is that even the games with a heavy amount of raytracing are still mostly using traditional rasterization methods. So if you can get even ~70% of the raytracing performance, but ~30% better rasterization overall, that’s a good deal.

And having used the 7900 GRE for 4k gaming for a few months now, even trying some light RT on it from time to time, if you gave me the same thing but two steps better on its main selling point, 4 steps better on the RT side, and a serious DLSS competitor on the software side, for the same price, I will happily stay right here on Team Red.

I’m feeling like it’s 2016 again with these rumors. I loved that old RX 480 8GB.

Summing up CES

Nvidia: AIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIAIA…

AMD: *kicking Intel while they’re down while pretending Nvidia isn’t bullying them*

Intel: *hanging on for dear life, but hey they made some decent budget GPUs this time*

The monitor industry: “700 frames, huh?… Yeah that’s cool and all. But, uh… *wiggles brow* Ours can hit 800.”

---

And that’s all for this issue of the Tech Update. If I didn’t mention it here, it clearly wasn’t important.

Will I do this next month? Who can say.