Hi, this is a quick post I make to talk about something I keep seeing online everywhere.
A lot of people say that the price increases will force developers to optimize and to work with what hardware they have to make good games and stop using AI gen and DLSS tech as an excuse for poor optimization.
The big problem is that nobody thinks about those people that don’t have the hardware right now.
Those people that were waiting for a discount to buy a PS5 or a PC and now they’re left stranded.
Current-gen consoles are getting really hard to find and a lot of people have been left out, stuck on old-gen and old-games.
playing old-games is not a bad thing but you may have missed the fact that even old consoles are getting reaaally pricey thanks to scalpers and speculators of the market.
This is madness people. Fight AI, don’t embrace it!
The devs won’t optimize anything. Our problems are not their problems.
Optimization happens when people have time to sit down and reason about the code. A hardware crunch isn’t going to give people more time. The assembly line is going to run at the same pace regardless, and devs are going to churn out the same junk they always have.
It isn’t a lack of limited hardware holding them back. Old, junky hardware has existed since nearly the beginning of the information technology, so a hardware crunch isn’t going to affect anything.
The big problem is that nobody thinks about those people that don’t have the hardware right now.
Literally yes they do, because even though they don’t have the latest and greatest hardware, they have some money to spend. That’s the argument being made: until now the assumption was that new hardware would get cheaper over time, and people would gradually move to new hardware. Devs spend years making games, and historically bank on that assumption so that when the game comes out, it has the largest audience available to purchase it.
The fact that it looks like that won’t be the case in the near future means devs have to shift their behavior to accommodate what their playerbase has, i.e. continue developing and optimizing the same hardware.
That said, this is all temporary. Whether they widen the pipeline, or the AI bubble bursts, in 2-3 years there will be a deluge of hardware hitting markets. (Provided trade/actual wars don’t get in the way, which is the bigger concern imo).
While this is mainly correct, I find it hard to believe the deluge of hardware is going to help us much. The current ones are AI specialized hardware, and switching to consumer hardware requires switching production for consumer hardware after the fact that the bubble pops. Supply will slowly fill the delayed demand.
They’re mostly not AI specialized, though. That’s why they’re so inefficient and why their demand contends with consumer hardware in the first place. Which makes sense, because AI is still in rapid development. They don’t know what the right answer is yet, but they know they need a bunch of fast memory and parallel processing.
The AI specific hardware being added to GPUs is still pretty general. CUDA cores are just parallel compute. Tensor cores are for doing parallel compute with fewer bits of precision. Yes, there are niche applications for fp16 and lower, but rendering is one of those applications.
We also need to accept that this isn’t the crypto bubble, this is the dotcom bubble. Like it or not, there is a real advancement in technology happening here, and it’s not going away. The bubble will pop because there’s far more money being invested per unit time than can be returned as profit per unit time, not because the tech is a farce. Yes, 99% of AI applications right now are a farce, but that 1% are giving us actual useful abilities we simply didn’t have before. Point being: our world after the bubble pops will still make use of AI, so any hardware over-production will still be useful to the general public for AI applications.
A lot of people say that the price increases will force developers to optimize and to work with what hardware they have to make good games and stop using AI gen and DLSS tech as an excuse for poor optimization.
The main problem with this is simply that it won’t happen. Every company would have to spend more on their games for no monetary gain, while their competition likely won’t do the same.
Not to mention, big game companies already have insanely powerful hardware, so it doesn’t impede the development in the slightest.
Im coming at it conceptually rather than a place of actual industry knowledge, but games of yesteryear were incredibly clever, pulling off all sorts of crafty hacky bullshit to get their games looking good on pretty low-end basic gear.
I’m not sure if that sort of thing is feasible or possible today, but I’m hoping it’ll push some of the bigger brains to find out. Necessity is the mother of invention and all that.
Of course it’s possible. Just not financially viable, according to corporate logic. It’s not about profit. It’s about ever increasing profit. So what do you do when the sales have reached their peak and stopped increasing? Lower the production costs and/or increase prices.
I’ve worked in the video game industry for a few years, both at and with large corporations as well as smaller studios. Game optimization has rarely, if ever, been a concern for anyone. Usually, as long as the fps only occasionally drop to 25 on high-end systems - it’s good enough.
To be clear, smaller game studios care significantly more about optimization/accessibility. There’s no denying that. However, with their limited resources, sometimes there’s not a whole lot they can do.
What you’re asking for is completely reasonable and would be great. But it’s just not gonna happen. Most studios prefer Unreal, because it lets them outsource a lot of work to India, and potentially cuts the development time ever so slightly.
the way games are going now is the way games have BEEN going. capitalism’s only purpose is infinite growth at any and all expense. this is what that looks like for gaming, bigger bigger bigger while bandaging it up with TAA and framegen and DLSS. when this house of cards tumbles, i know i will capitalise on it by not investing in the same mistakes again. but, i can only really hope that others do the same.
I was gonna upgrade when the steam machine came out but with the likely price hike taking it to $800+ I think I’ll just keep playing 20 year old Call of Duty: United Offensive
Pretty sure it was around 800 before the price hike. They said they wouldn’t subsidize the costs and the parts specified would have put them close to 800
I thought the steam deck would have been priced higher than it was as well, so I thought that they’d probably surprise us with a $600 machine. $800 made sense for the specs (Now I expect it to be $1000)
Once GPU prices super inflated 6 or 7 years ago, I got out of PC gaming for new titles. I built a top spec machine for $1200 in 2012 (when I had disposable income and no kids), that barely gets you an entry level today. I guess PC gaming, for new titles at least, will forever be out of my budget.
800 is a fair price. Similarly specced minipc HX99G was sold at $700-900 until it went out of stock. I am beginning to doubt that Steam Machine is goint to be sold at this price. Must be higher due to ram and ssd prices of today. In other case, 800 for Steam Machine would be a steal, really.
I view it less as a defense of the current situation and more people trying to find what little optimistic takes are left to take while dealing with this hobby killer greed.
I don’t embrace it, but I can do nothing about it. I’m just going to try to keep what hardware I have alive as long as possible with DIY repairs where I can and buy overpriced used parts when I must including oddball stuff from China.
At the moment, my NAS is becoming unaffordable to maintain if things stay as they are. Datahoarders weep.
They can just buy used shit and not play at 4k resolution with 120fps. That stuff will also last longer now.
I just spent a few hundred to upgrade my gaming computer. I bought an old amd r5 5600x from 2020 for $150 and an old amd rx 6600 xt that came out 5 years ago.
I’m more than happy to let the “ever more demanding giant game” trend disappear. Half the games I play use pixel art graphics anyhow. If this pushes devs to make smaller and more creative games that run on decade old systems to reach the most gamers, I’m fine with this, still. Less e waste. Less constant feeling of needing newer components. Less bloatware from shitty games that want 100+GB storage.
Before prices get outrageous on 5+ year old components all these bullshit AI and data centers buying up everything will be dumping the current high end stuff out on the market to make way for newer stuff. And that’s IF the ai bubble doesn’t burst beforehand.
Used parts will also command a higher price, and also who is selling their old rig if hardly anyone can afford an upgrade?
I’m 90% certain that none of the resources being “bought” by AI actually have been sold, or created yet.
Either warehouses of GPUs and RAM are gonna go in a shredder in 12 months, or they’re just decreasing supply and charging more, like the auto industry.
Speaking of the auto industry: Is AI a mirage, like the dreams of a working rotary engine? Many companies have tried, but it keeps killing companies and it still is an impossible goal. The technology projects a mirage for investors that it just can’t reach.
AGI isn’t coming out of LLMs and statistical weights.
Their model, which has scraped the internet and therefore by definition knows how to make a pipe bomb, cannot be proven that it wont tell the user how.
The guard rails are impossible to build because LLMs aren’t deterministic.
I’m 90% certain that none of the resources being “bought” by AI actually have been sold, or created yet.
Indeed. The prices skyrocketed because vendors realised they couldn’t get replacement supply in the future. What existed today was all they were going to get.
I’m expecting a glut of supply once those contracts fall through.
Speaking of the auto industry: Is AI a mirage, like the dreams of a working rotary engine?
It is, but I think it’s a different type of mirage. The rotary engine does work, but it brings with it significant downsides. Getting the positives without the negatives is the mirage being chased.
AI appears to do one thing, but actually does another. People see it “creating” new things, but it’s more like it shreds work up and then glues the pieces together making sure it looks consistent. Train it on one work and it can reproduce that work. Train it on two and it will mash the two. Train it on a billion and it will mash the billion. Nothing creative,. No extrapolation. Just interpolation.
People want the AI promise regardless of the downsides. It just doesn’t exist.
The mirage is consciousness or AGI out of the plagarism machine.
like the dreams of a working rotary engine? Many companies have tried, but it keeps killing companies and it still is an impossible goal.
Rotarys and 2-strokes are kneecapped by emissions standards/laws, not because they don’t work.
Gyroscopic effect: “Am I just a joke to you?”
Rotary seals go brr.
If that’s the case where are they? Other than being used in israeli suicide drones.
Motorsports. Altho emission/pollution regulations are pushing them out of that too.
Dozens of annual sales I’m sure.
You’re mixing up “miracle tech that leads to nowhere” with “niche tech with little mass appeal”. A rotary engine car has won Le Mans, The Mazda 787. I’m pretty sure one of the recent Mazda plug in Hybrids (I refuse to call those EV-s) has a rotary engine as a backup for the electric engine.
I’m truly sorry 2-smokers peed in your cereal.
But you’re proving my point. They’re a dead end and technologically LLM as a vector towards AGI is like thinking Rotaries are a vector for motoring… They’re not and they’re not.
👍
Shit‘s been fucking expensive before this year and most triple-A stuff ran like doodoo anyway. Now prices are just even worse and somehow that‘ll make the suits care? At best we‘ll get the same level of bad.
MH Wilds has been out for a year and today‘s patch pushed it to the bottom range of acceptable.
Even if somehow more money was put into engineering now, the argument would still be weak, it‘s naive on its own merit at best.
Honestly I hope the ENTIRE console industry completely dies off. Hopefully Nintendo bites it first, but they’re all fucking shitty as hell (for SO many reasons) and I hope they all go extinct by the time Trump does.
There, I said it, I’m not sorry, and I will die on this hill. I don’t even think there’s any reasonable counter point beyond it being a simple entry point with easy to plug in pre-configured boxes.
So fight me. Consoles suck, and they should go extinct.
I don‘t really see a future where game consoles die but gaming PCs don‘t because of hardware shortages. It‘s either cloud all the way or this becomes the era of mobile gaming even for core gamers.
Personally I hope we can somewhat return to normal in a few years. That is after the bubble popped and even the last investor realized most data centers won‘t get built anymore.
Which would you rather have as the dominant platform. Consoles, or cloud gaming?
Because if “market conditions” kill consoles, they will shrink PC gaming hardware sales too, and I don’t want a world where devs target cloud gaming first.
I’m not trying to defend consoles and their predatory practices, but you can’t separate them out. If subsidized console hardware is too pricey to sell, then PC gaming components will absolutely atrophy too.
First, I wasn’t saying it would be market conditions doing it. Honestly I’d love to see all the companies eat the shit they’ve already stepped in.
Preferably it would be people wising up and realizing that they are factually bad compared to PCs. Demand would (in this good timeline) drop to zero overnight and kill them off immediately.
Not sure why you’re so sure that cloud would be the next winner either. Until network speeds get above the speed of light, as long as real time games exist, cloud gaming will never be very popular. It’ll be at best the “gaming you have at home” meme.
The delay will always be too much for any serious game where real time input and reactions are core components.
So I don’t see cloud gaming ever getting huge.
And even in your version, if PCs do take a hit (they already are with the RAMpocalypse), it’s still a smaller hit than being defunct.
But overall the main point is no technology should ever be locked down to one company. If your hardware only plays games allowed by one company, then you’ve got yourself a piece of shit. Period.
Not sure why you’re so sure that cloud would be the next winner either.
Because, in aggregate, gamers are stupid consumers.
I hate to be so blunt, but they have, repeatedly and demonstrably, made uninformed purchases. They buy bad games on launch day, complain, then turn around and do it again. They buy hardware known to be a lemon. Heck, they’ll hardly even look at AMD or Intel GPUs now simply because there’s isn’t a minimum amount of effort made to shop around.
They are going to just buy the cloud gaming subscriptions if that’s all that’s financially viable, and it’s what’s popular in their YouTube feeds or Discord channels or whatever.
Keep in mind that I’m talking about the bulk market. Sure, plenty of us will turn our nose up. But the R&D required to develop consumer hardware requires volume, so updates will get slimmer with less money in the pool. It’s already happened with the AMD 9000 GPUs (as shrinking sales could not justify a big-die 7900 successor).
The average gamer and consumer will never wise up, especially with social media keeping them stupid.
Tbh all over simplified tech going away would be nice. Force people to learn and think to use devices and services, at least get away from the “app and console” type of mindsets.
Yours is an aggressive timeline, but I think the market is naturally trending that way for a lot of reasons.
Challenge accepted!
The Steam deck is pretty cool because it is just a really nice handheld PC that lets me (or, more typically, my wife) play everything in my very extensive Steam library.
Maybe it was unclear, but by “console” I was implying a locked down device stuck with only using software licensed (or whatever the applicable legal term is) to a specific company’s hardware.
And given that a steam deck (love mine btw) isn’t locked down in any way beyond having native access to a specific vendor’s store, doesn’t apply (neither does steam machine).
Any locked down technology at all is kinda suspect, tbh. Capitalism is fucking horrendous.
Well met, sir!
That’s a weird take.
I still rock a gtx 1060, I have no issue playing a wide variety of games, obviously most classics and many newer indie titles.
The games I “can’t run” are modern AAA titles that put a lot of emphasis on spectacle and pay no attention to optimization.
Yes it sucks for people who want new hardware right now because they have literally nothing, but even then something used from 5-10 years ago will play 95%+ of all games, including many many classics and very popular games like minecraft and fortnite.
Problem in that many games using UE5 are coming out, even AA and indy games. A GTX 1060 usually won’t do for UE5 unless you accept severely degraded graphics.
UE5 has been getting flak this year for being a notoriously unoptimized mess. Any decent game designer should be avoiding them by this point.
AA and indy games. A GTX 1060 usually won’t do for UE5 unless you accept severely degraded graphics.
- well yes, I do.
- I don’t usually play games like that, I think, you are free to name a few you think this is an issue for.
- indie art isn’t that “high cost of investment/valuable” anyway. Meaning, they don’t have 500 people creating high vertex count 4k textures everywhere. YAGNI, I don’t believe this is an issue in practice.
- UE is a commercial engine. With support and dev staff. And subscription pricing and everything. They can optimize?!
- Yes indies need to use the optimizations or build them themselves. Skill issue.
I’m not calling you wrong, I doubt I could play “expedition 33” in “nice graphics”, but I have 0 interest in JRPGs, so it’s literally not a problem for me.
I’d say we should just stop buying these games, but it seems people have no standards anymore and will just buy anything marketed to them. Even if its a 30fps obvious cash grab.
Modern developers simply were not ever taught the old ways. The old ways are light years ahead in terms of optimization and efficiency than what modern developers can even imagine. Imagine creating a C compiler and it has to run in 16KB of RAM and be powerful enough to build unxz, untar and sha256sum. Because it was done because they needed it.
They also aren’t given enough time to even do optimization
They’re also not really comparable. Teams were so small and project timelines were so short that you often knew exactly what the end would look like. My favorite optimization story from 20+ years ago is that a dev (who went nameless, and so did the game, as the story was posted anonymously) made a habit of declaring a large empty variable at the beginning of a project, and that variable’s only job was to be deleted when they encroached on their memory budget so they knew when to stop.
Do you have an example or a snippet in mind for a piece of modern game engine code that would have been done better in the “old ways”?
That someone thought Electron was a not just a reasonable approach but a good idea, when it sucks down a gig of RAM for what amounts to mIRC with GIFs, is a strong support of your claim.
Optimizing for development time is a worthy pursuit as well.
Maybe developers could use that oh-so-amazing AI to optimize then…
mIRC with gifs? Some people thought it was a good idea to use it as a game engine! It was used for e.g. Shapez
Slack was one of the earliest Electron apps. Then Teams.
They pay for dev time, the extra energy costs and worse experience to the user are t in their budget.
They all work for big companies, and in big companies you get paid for features not for fixes.
AAA games will never be optimized. They’re too big to care, and nobody is accountable for it.
This is why I quit big companies. I love optimizing. I love building flexible software, that’s fast, clean, and simple. But that takes time and you won’t get it, you’ll get “what’s the minimum we can do to get this feature out” or “we can always come back to that” and it’s back logged forever. If someone else pushes out a hacky MVP of spaghetti code that gets to market faster, they’ll go with that, even if it costs months of dedicated fixes.
While I understand the pain of someone having no hardware right now, I think you should still be able to get an old gaming PC or an old Playstation 4 to enjoy games for cheap.
In fact, almost everyone has something like this lying around…
While it means running old AAA or lower demanding current games, it’s still great gaming time.
It’s not just an AI problem.
You’ve pointed out the scalpers and speculators of the market. Are we going to do anything about them? No? It’s all just AI, right?
What good is it for the developers when they’re whipped by companies with unrealistic demands and people who’re strained from paying top-dollar for the games? It’s all just AI!
It’s a multi-pronged issue.
We know, but the one good thing AI has done is help bring these other issues to light.
However, having no supply of hardware because it’s all been “preordered” is a much larger issue than scalpers.
Some companies have protections against scalping, some dont care, but the biggest issue there is people buying the “scalped” products. They just cant help themselves.
Developers are forming more unions, its picking up pace, so there’s hope there. It’s also partly a customer issue too, buy and support the games made ethically.
